IN THE SUPREME COURT OF
CALIFORNIA
GILEAD TENOFOVIR CASES.
S283862
First Appellate District, Division Four
A165558
San Francisco City and County Superior Court
JCCP No. 5043
August 3, 2026
Justice Groban authored the opinion of the Court, in which
Justices Corrigan, Liu, Kruger, and Desautels* concurred.
Chief Justice Guerrero filed a concurring opinion.
Justice Kruger filed a concurring opinion in which Justices
Corrigan and Desautels* concurred.
Justice Evans filed a dissenting opinion.
________________________
* Associate Justice of the Court of Appeal, First Appellate
District, Division Two, assigned by the Chief Justice pursuant to
article VI, section 6 of the California Constitution.
GILEAD TENOFOVIR CASES
S283862
Opinion of the Court by Groban, J.
This case requires us to determine whether a drug
manufacturer may be liable in negligence for injuries allegedly
caused by a drug used by plaintiffs that plaintiffs concede is not
defective. Specifically, plaintiffs allege that defendant Gilead
Sciences, Inc. (Gilead), is liable in negligence because it
unreasonably delayed bringing to market another drug it had
invented that was equally effective and less toxic than the
nondefective drug that allegedly caused plaintiffs’ injuries.
Plaintiffs allege that they would have switched to the
alternative drug and avoided injury had Gilead not
unreasonably delayed bringing the alternative drug to market.
The Court of Appeal held that liability may attach in these
circumstances. (Gilead Tenofovir Cases (2024) 98 Cal.App.5th
911, 916–917.)
The Court of Appeal acknowledged our precedent
suggesting that, when a plaintiff sues a manufacturer under a
products liability theory for injuries caused by one of its
products, the plaintiff generally must prove that the product
was defective. (See Gilead Tenofovir Cases, supra,
98 Cal.App.5th at p. 929.) But the court nonetheless considered
it an open question whether alleging a defect is always
necessary, and concluded plaintiffs need not allege a defect here.
(Id. at pp. 929–930.) The Court of Appeal reasoned that
plaintiffs need not allege a defect because they could proceed
instead under a general negligence theory pursuant to Civil
1
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
Code section 1714, subdivision (a) (section 1714), which
establishes a default duty requiring all persons and entities to
exercise reasonable care to avoid causing harm to others. In
adopting that view, the court rejected Gilead’s contention that a
drug manufacturer’s only duty under our precedent is the duty
to market products free from defects. (Gilead Tenofovir Cases,
at pp. 922, 934.)
Although the Court of Appeal correctly explained that
section 1714 establishes the baseline duty to exercise reasonable
care to avoid injuring others, what constitutes reasonable care
in a particular context may be defined by statutes, regulations,
or judicial decisions. (Parsons v. Crown Disposal Co. (1997)
15 Cal.4th 456, 461 (Parsons); Ramirez v. Plough, Inc. (1993)
6 Cal.4th 539, 547.) This case concerns a drug manufacturer’s
duty to prevent harm arising from the use of its products.
Decades of products liability precedent have defined that duty
as limited to designing, manufacturing, and marketing products
that are free from defects. (Trejo v. Johnson & Johnson (2017)
13 Cal.App.5th 110, 125 (Trejo); Merrill v. Navegar, Inc. (2001)
26 Cal.4th 465, 478–479 (Merrill).) That precedent lends
support for Gilead’s view that a manufacturer’s duty of
reasonable care under section 1714 is limited to designing,
manufacturing, and marketing products that are free from
defects, such that a manufacturer cannot be held liable for
injuries caused by a nondefective product.
Moreover, holding that manufacturers owe a broader duty
to not unduly delay developing and commercializing an
allegedly safer product to replace a concededly nondefective one
could conceivably upend current products liability law. Instead
of having to prove that a product is defective — an essential
element of any products liability claim, whether sounding in
2
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
negligence or strict liability — plaintiffs would need only to
convince a fact finder that the manufacturer acted unreasonably
in its development and commercialization decisions. This
theory of negligence liability lacks a clear limiting principle and
raises serious questions about whether a workable standard for
assessing the reasonableness of such decisions could ever be
established. It also risks inviting fact finders to second-guess
complex resource-allocation decisions about whether and when
to pursue potential alternative products while a concededly
nondefective product remains on the market. The prospect of
second-guessing presents particularly acute concerns in the
prescription drug context, where safety and efficacy
assessments are often provisional and uncertain during early
stages of clinical research.
We need not, however, definitively determine whether a
drug manufacturer may ever be liable in tort for negligent
conduct that results in injuries from taking a nondefective drug.
Even if we were to assume arguendo that manufacturers may
owe a general duty of reasonable care apart from their duty to
market products free from defects, we held in Rowland v.
Christian (1968) 69 Cal.2d 108 (Rowland) that foreseeability
and policy considerations may justify an exception to section
1714’s default duty in appropriate circumstances. Such
circumstances exist here.
Where, as here, the allegedly safer drug has not yet
undergone large-scale clinical testing in humans or received
approval from the federal Food and Drug Administration (FDA),
any harm resulting from a drug manufacturer’s delay in
commercializing that drug would arise, if at all, only through a
chain of uncertain scientific outcomes and discretionary
decisions by actors beyond the manufacturer’s control.
3
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
Imposing a duty of care in these circumstances would place
extraordinary burdens on drug manufacturers by effectively
requiring them to commit substantial time, expenses, and
resources to conduct the later-stage clinical trials necessary to
obtain FDA approval. It would also risk distorting research
priorities and chilling pharmaceutical innovation in ways that
may ultimately undermine, rather than advance, public health
and safety.
We accordingly hold that, even assuming drug
manufacturers owe a broader duty of care apart from their duty
to design, manufacture, and market nondefective drugs, the
Rowland factors compel an exception to such a duty.
I. BACKGROUND
Before providing the factual background of Gilead’s
development of the drugs at issue, it is helpful to briefly outline
the regulatory framework governing the testing and approval of
new prescription drugs. Under the Federal Food, Drug, and
Cosmetic Act (FDCA) (21 U.S.C. § 301 et seq.), a drug
manufacturer must gain approval from the FDA before
marketing any new drug. New drugs are initially studied in the
laboratory and tested in animals to evaluate toxicity and
determine whether they can be safely administered to humans
in clinical trials. (See 21 C.F.R. § 312.23(a)(8) (2022); FDA, The
FDA’s Drug Review Process: Ensuring Drugs Are Safe and
Effective (Nov. 24, 2017) <https://www.fda.gov/drugs
/information-consumers-and-patients-drugs/fdas-drug-review-
process-ensuring-drugs-are-safe-and-effective> [as of August 3,
4
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
2026] (hereafter Drug Review Process).)1 If these preclinical
studies yield favorable results, the manufacturer may submit an
investigational new drug application (IND) with the FDA to
obtain authorization to begin human clinical trials. (See
generally 21 C.F.R. § 312.1 et seq. (2022); see also Drug Review
Process, supra.)
Human testing generally proceeds in three phases. Phase
I trials typically involve administering a new drug to a small
group of healthy volunteers or patients with the targeted
condition — typically between 20 and 80 individuals — to collect
preliminary data on safety, tolerability, and appropriate dosage.
(21 C.F.R. § 312.21(a) (2022); FDA, The Drug Development
Process, Step 3: Clinical Research (Jan. 4, 2018) <https://
www.fda.gov/patients/drug-development-process/step-3-
clinical-research> [as of August 3, 2026] (hereafter Step 3:
Clinical Research).) Phase I trials are intended “to identify
rudimentary product characteristics, such as how the body
metabolizes a drug and how long it stays in the body, and to
provide evidence that the product is not too toxic for further
human testing.” (FDA, 22 Case Studies Where Phase 2 and
Phase 3 Trials had Divergent Results (Jan. 2017) p. 2 (hereafter
22 Case Studies) <https://www.fda.gov/media/102332/
download> [as of August 3, 2026].) The FDA estimates that
approximately 70 percent of drugs tested in phase I advance to
phase II. (Step 3: Clinical Research, supra.)
1
All Internet citations in this opinion are archived by year,
docket number, and case name at <https://courts.ca.gov/
opinions/cited-supreme-court-opinions>.
5
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
Phase II trials test the drug on up to several hundred
subjects who have the targeted condition, with the goal of
evaluating efficacy and identifying common short-term side
effects and risks. (21 C.F.R. § 312.21(b) (2022); Step 3: Clinical
Research, supra.) These trials “are intended to explore the
effectiveness of the product for a particular indication over a
range of doses, and to assess short-term side effects.” (22 Case
Studies, supra, at p. 2.) They “often measure laboratory values
or other biomarkers rather than clinical outcomes (i.e., effects
on how a patient feels, functions, or survives).” (Ibid.) When
clinical outcomes are assessed in phase II trials, “it is usually
for relatively short periods of time and in a relatively small
number of people.” (Id. at p. 3.) Only about 33 percent of drugs
tested in phase II advance to phase III. (Step 3: Clinical
Research, supra.)
Phase III trials enroll several hundred to several thousand
subjects and are conducted over one to four years to obtain the
additional information on safety and efficacy necessary to
evaluate the drug’s overall benefit-risk profile. (21 C.F.R.
§ 312.21(c) (2022); Step 3: Clinical Research, supra.) These
trials “generally assess clinical outcomes, and are designed to
determine whether the demonstrated benefits of the product
outweigh its risks.” (22 Case Studies, supra, at p. 3.) Because
phase III studies are larger and longer in duration, they
“provide most of the safety data” and are more likely to reveal
long-term or rare side effects. (Step 3: Clinical Research, supra.)
Even at this stage, only approximately 25 to 30 percent of drugs
will be considered successful enough to present results to the
FDA for approval. (Ibid.) After completing clinical trials, drug
manufacturers may file a New Drug Application (NDA) with the
FDA and, if approved, may begin marketing the new drug. (See
6
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
generally 21 C.F.R. § 314.1 et seq. (2022); see also 21 C.F.R.
§ 314.105(a) (2022); Drug Review Process, supra.)
Gilead followed this regulatory framework in obtaining
approval to market tenofovir disoproxil fumarate (TDF) and
tenofovir alafenamide fumarate (TAF), both of which are used
to treat human immunodeficiency virus (HIV). In 1991, Gilead
obtained the exclusive right to develop, manufacture, and sell
tenofovir-based medications for the treatment of HIV. Prior to
this time, HIV had a 95 percent fatality rate. Tenofovir-based
medications such as TDF and TAF would eventually become a
cornerstone of HIV antiretroviral therapies, which save millions
of lives worldwide each year, as well as a staple in pre-exposure
prophylaxis therapies used to prevent HIV infections. (See
Ustianowski & Arends, Tenofovir: What We Have Learnt After
7.5 Million Person-Years of Use, Infectious Diseases and
Therapy (June 2, 2015) vol. 4, pp. 145–146,
<https://link.springer.com/article/10.1007/s40121-015-0070-1>
[as of August 3, 2026]; FDA, The History of FDA’s Role in
Preventing the Spread of HIV/AIDS (Mar. 14, 2019)
<https://www.fda.gov/about-fda/fda-history-exhibits/history-
fdas-role-preventing-spread-hivaids> [as of August 3, 2026].)
In March of 1997, after conducting preclinical research on
TDF, Gilead filed an IND with the FDA to obtain approval for
human clinical testing of its first TDF-based medication, which
would eventually be called Viread. After four years of clinical
research of TDF’s safety in humans, Gilead filed an NDA for
Viread in April 2001, which the FDA approved in October 2001.
Gilead subsequently obtained FDA approval to market other
TDF-based medications.
7
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
Around the same time that it was conducting clinical
testing on TDF, Gilead began investigating TAF as a potential
backup to TDF. In November 2001, just one month after
obtaining FDA approval to market Viread, Gilead filed an IND
to begin human clinical testing of TAF. Gilead then conducted
a 14-day combined phase I/II clinical trial with 30 subjects — 20
receiving TAF and 10 receiving TDF — to make initial
comparisons of the safety and effectiveness of TAF relative to
TDF. Plaintiffs contend that this study confirmed what Gilead
had already learned in its preclinical research: TAF could
achieve the same antiviral effect as TDF at a significantly lower
dosage, meaning it was as effective as TDF but with
considerably reduced risks of renal, bone, and tooth injuries.
Plaintiffs allege that, even though Gilead was aware that TAF
was safer than TDF, it announced in 2004 that it had ceased
developing TAF because its safety, tolerability, and efficacy
profile did not appear sufficiently different from TDF. Plaintiffs
claim that this explanation was pretextual and that Gilead
discontinued TAF development because it believed TAF would
cannibalize TDF’s sales. According to plaintiffs, Gilead chose to
wait until TDF’s patent expired before bringing TAF to market
in order to maximize profits on both drugs.
Gilead disputes that it knew TAF was safer than TDF
when it chose to discontinue TAF development, or even that it
could have known TAF was safer prior to conducting phase III
clinical trials. It also challenges plaintiffs’ assertion that it
delayed TAF development solely to maximize profits. After the
Court of Appeal requested supplemental briefing on Gilead’s
knowledge about TAF’s safety relative to TDF, Gilead argued
before this court that, had TAF shown “meaningful
improvement” over TDF in the combined phase I/II clinical trial,
8
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
it would have gone “all-in on TAF” because, although it expected
TAF to “ ‘cannibalize’ TDF’s market share,” it also anticipated
that TAF could attract new patients if superior to TDF, yielding
“an extra $1 billion over expected TDF revenues between 2008
and 2013.” But because TAF did not demonstrate superiority to
TDF, Gilead saw no reason to rush it to market: TAF “would
only ‘cannibalize Viread’ without helping current patients or
attracting new ones,” and Gilead “would see no added revenue.”
Gilead asserts that it later resumed TAF development “to
address a new problem”: “People were living longer with HIV
than anyone dared to hope, and with age comes bone-density
loss and reduced kidney function, the same rare side effects
associated with TDF.” It decided to further explore TAF “on the
chance that it might prove to be a lower-dose alternative for that
aging population,” though its researchers “expressed continued
uncertainty that TAF was any safer than TDF.”
Plaintiffs concede that the preclinical and clinical data
Gilead possessed on TAF in 2004 — including the results of the
phase I/II clinical trial — were insufficient to obtain FDA
approval or bring TAF to market. Plaintiffs nonetheless claim
that Gilead unreasonably delayed taking the steps necessary to
commercialize TAF, including conducting the larger-scale
human clinical trials required for FDA approval, and that this
decision breached a duty of care to them.
Gilead resumed TAF development in 2010. It conducted
its first phase III clinical trial on TAF in 2013, and it ultimately
obtained FDA approval for TAF in November 2015. Plaintiffs
assert that this timing was deliberate, maintaining that Gilead
waited until 2010 to resume TAF development to ensure that
the drug reached the market in 2015 — just ahead of the
expiration of Gilead’s TDF patent in 2017. According to
9
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
plaintiffs, this strategy allowed Gilead to promote TAF as a
safer, improved alternative, preserving a competitive advantage
against generic versions of its TDF-based medications.
TDF-based medications remained on the market when
Gilead commercialized TAF in 2015, and they continue to be sold
today. Plaintiffs do not contend that TDF is defective or that
Gilead should have withdrawn it from the market. Indeed,
plaintiffs acknowledged in the Court of Appeal that some
patients prefer TDF over TAF. Instead, plaintiffs assert that
Gilead unreasonably delayed bringing TAF to market — which
they believe could have been commercialized as early as 2006 —
and that this delay deprived them, for many years, of the option
to take what they allege is a safer drug. Plaintiffs allege that,
as a result, they suffered serious renal, bone, or tooth injuries
from taking TDF that could have been avoided had TAF been
made available sooner.
At the time Gilead moved for summary judgment, only two
claims remained: negligence and fraudulent concealment. The
trial court denied Gilead’s motion. Gilead then petitioned for
writ of mandate, and the Court of Appeal issued an order to
show cause. (Gilead Tenofovir Cases, supra, 98 Cal.App.5th at
p. 917.) The court ultimately granted the petition in part and
denied it in part, directing the trial court to grant summary
adjudication as to the fraudulent concealment claim but leaving
intact the trial court’s denial of summary adjudication on the
negligence claim. (Ibid.) The court reasoned that “the legal duty
of a manufacturer to exercise reasonable care can, in
appropriate circumstances, extend beyond the duty not to
market a defective product.” (Ibid.) The court further concluded
that, on the record before it, the Rowland factors did not support
10
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
recognizing an exception to the general duty of care. (Gilead
Tenofovir Cases, at p. 917.)
We granted review.
II. DISCUSSION
A. Standard of review
In reviewing a denial of summary judgment, we usually
apply a de novo standard of review and consider “ ‘ “all the
evidence set forth in the moving and opposing papers except that
to which objections were made and sustained.” ’ ” (Yanowitz v.
L’Oreal USA, Inc. (2005) 36 Cal.4th 1028, 1037.) Here, however,
the Court of Appeal accepted the allegations of the complaint as
true on certain issues — including Gilead’s alleged knowledge
that TAF was safer than TDF and its profit motivations —
reasoning that Gilead “did not contest those assertions for the
purposes of the summary judgment motion presently under
review.” (Gilead Tenofovir Cases, supra, 98 Cal.App.5th at
pp. 921–922; see also American Airlines, Inc. v. County of San
Mateo (1996) 12 Cal.4th 1110, 1118 [“When a motion for
summary judgment is used to test whether the complaint states
a cause of action, the court will apply the rule applicable to
demurrers and accept the allegations of the complaint as true”].)
Gilead contends that the Court of Appeal applied the wrong
standard of review and that the court should have considered
evidence it believes undermines plaintiffs’ assertions regarding
its knowledge and motivations.
Although Gilead assumed the truth of plaintiffs’
allegations for purposes of summary judgment, that assumption
must be understood in context. Gilead moved for summary
judgment solely on the ground that a negligence claim based on
injuries caused by a nondefective product is not cognizable. The
11
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
question Gilead believed it was litigating — whether a plaintiff
must prove that the injury-causing product was defective — was
a pure question of law to which Gilead’s knowledge regarding
the safety and efficacy of TAF was irrelevant.
The issue changed before the Court of Appeal. After oral
argument, the court sua sponte requested supplemental briefing
on several issues, including whether plaintiffs contend that
Gilead knew or reasonably should have known that TAF was a
safer alternative, what plaintiffs must prove Gilead knew or
reasonably should have known about TAF relative to TDF to
establish a duty to continue developing TAF, and whether the
Rowland factors require establishing an exception to any such
duty. In its eventual opinion, the court rejected Gilead’s
argument that a drug manufacturer cannot be held liable for
injuries resulting from a nondefective product and held that a
manufacturer owes a general duty of reasonable care to users of
a nondefective drug when deciding whether and when to
commercialize an alternative drug it allegedly knows is safer
and equally effective. (Gilead Tenofovir Cases, supra,
98 Cal.App.5th at p. 922.) Once the Court of Appeal framed the
duty in those terms, knowledge became central to the question
of liability.
Given that shift, and because the Rowland analysis must
be conducted at a “ ‘categorical’ ” rather than a “ ‘case-specific’ ”
level (Kuciemba v. Victory Woodworks, Inc. (2023) 14 Cal.5th
993, 1021 (Kuciemba)), we find it appropriate to consider
whether any drug manufacturer could reliably know during the
early stages of clinical testing that a drug in development is
safer than an existing one.
12
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
B. Negligence without proof of a defect
In the products liability context, it is well settled that
manufacturers have a duty to market products free from defects
in design, manufacture, or labeling. (Trejo, supra,
13 Cal.App.5th at p. 125.) For products other than prescription
drugs, California law eases a plaintiff’s burden by imposing
strict liability for design defects without requiring proof of
negligence. (Barker v. Lull Engineering Co. (1978) 20 Cal.3d
413, 431 (Barker).) A plaintiff may alternatively bring a claim
for negligent design, manufacture, or labeling of a product
(Jiminez v. Sears, Roebuck & Co. (1971) 4 Cal.3d 379, 387), and
must do so when alleging a prescription drug is defectively
designed (Brown v. Superior Court (1988) 44 Cal.3d 1049, 1061,
1065 (Brown)), but such claims still require proof that the
manufacturer’s negligence resulted in a defective product
(Jiminez, at p. 383; Merrill, supra, 26 Cal.4th at pp. 478–479).
Against this backdrop, Gilead argues that the product
defect requirement limits the scope of a drug manufacturer’s
default duty of reasonable care established by section 1714, such
that a manufacturer cannot be held liable in negligence for
injuries caused by a nondefective product. Plaintiffs, by
contrast, contend that section 1714 imposes liability for
negligent conduct that creates a risk of harm even when the
injury-causing product is not defective. The Court of Appeal
agreed with plaintiffs. For the reasons discussed below, we have
significant doubts regarding plaintiffs’ theory of liability. But
even assuming plaintiffs’ theory is legally viable and that a
manufacturer may owe a duty of care under section 1714 for
13
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
injuries caused by a nondefective product, the Rowland factors
justify an exception to any such duty in this context.
Before turning to those factors, we offer some preliminary
observations as to why recognizing a duty untethered to a
product defect would be difficult to reconcile with settled
products liability principles. Plaintiffs are correct that, absent
a statutory or judicial exception, all persons and entities are
subject to section 1714’s rule that everyone is responsible for
injuries “occasioned to another by his or her want of ordinary
care or skill in the management of his or her property or person.”
(§ 1714, subd. (a).) In other words, persons and entities are
generally “ ‘ “liable for injuries caused by [their] failure to
exercise reasonable care in the circumstances” ’ ” of a given case.
(Cabral v. Ralphs Grocery Co. (2011) 51 Cal.4th 764, 771
(Cabral).) “[T]he law imposes a general duty of care on a
defendant only when it is the defendant who has ‘ “created a
risk” ’ of harm to the plaintiff, including [situations in which]
‘ “the defendant is responsible for making the plaintiff’s position
worse.” ’ ” (Brown v. USA Taekwondo (2021) 11 Cal.5th 204, 214
(USA Taekwondo).) The relevant inquiry is not whether the
defendant committed or omitted a particular act, but rather
“ ‘whether the actor’s entire conduct created a risk of harm.’ ”
(Id. at p. 215, fn. 6.)
We have long recognized, however, that this default duty
may be refined and limited in particular contexts through
judicial decisions defining the “proper conduct of a reasonable
person under particular situations.” (Satterlee v. Orange Glenn
School Dist. (1947) 29 Cal.2d 581, 587; accord, Kentucky Fried
Chicken of Cal., Inc. v. Superior Court (1997) 14 Cal.4th 814,
824.) In the products liability context, our decisions have
generally defined a manufacturer’s duty under section 1714 as
14
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
the duty to design, manufacture, and market products that are
free from defects. Recognizing a broader duty under which
liability may arise even absent a defect in the injury-producing
product would create substantial tension with that body of law.
To explain, products liability law employs specific tests to
determine whether a product is defective. Under the risk-
benefit test for design defects, the fact finder weighs the gravity
and likelihood of the danger posed by the design against the
feasibility, cost, and potential drawbacks of a safer alternative
design. (Barker, supra, 20 Cal.3d at pp. 430–431.) Likewise, a
plaintiff alleging a product defect stemming from a failure to
warn must establish that the manufacturer knew or had reason
to know the product was dangerous for its intended use, had no
reason to believe users would recognize the danger, and failed
to exercise reasonable care to provide an adequate warning.
(Stevens v. Parke, Davis & Co. (1973) 9 Cal.3d 51, 64.)
Regardless of the governing test, liability turns on whether the
product is defective. (Merrill, supra, 26 Cal.4th at pp. 479–480.)
Plaintiffs’ theory risks circumventing these settled
principles. If manufacturers owed a broader duty to act
reasonably to avoid all product-related injuries, plaintiffs could
seek recovery for harms caused by concededly nondefective
products merely by alleging the manufacturer acted
unreasonably. Questions currently governed by established
defect tests would instead be left to a fact finder’s generalized
assessment of reasonableness.
This problem is especially acute in the context of
prescription drug development. A generalized negligence
standard would invite fact finders to second-guess complex
research, development, and resource-allocation decisions made
15
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
under conditions of scientific uncertainty and incomplete
information, including decisions regarding whether and when to
pursue potential alternative drugs while a concededly
nondefective drug remains on the market. Unlike product defect
claims, which are evaluated based on contemporaneous and
relatively complete scientific evidence available at the time of
the product’s distribution (see Carlin v. Superior Court (1996)
13 Cal.4th 1104, 1116 (Carlin)), evaluating the reasonableness
of drug-development decisions would require retrospective
assessment of inherently provisional judgments made based on
evolving scientific data.
Perhaps recognizing these difficulties, the Court of Appeal
attempted to narrow the duty by limiting it to situations in
which a drug manufacturer has already invented or developed a
drug it knows to be safer and equally effective. (Gilead
Tenofovir Cases, supra, 98 Cal.App.5th at pp. 922, 944.)
Although the Court of Appeal purported to select the term
“ ‘invent’ ” rather than “ ‘develop,’ ” on the ground that “the
meaning of ‘develop’ in the pharmaceutical context is
ambiguous” (id. at p. 921, fn. 3), it later described the duty as
requiring a manufacturer to act with reasonable care when it
has “developed an alternative that it knows is safer and at least
equally efficacious” (id. at p. 944, italics added). Both terms are
ambiguous and, for that reason, are unlikely to act as
meaningful constraints in practice.
For example, what does it mean for a drug to be
“invented”? Does invention occur when the compound is first
conceived, or when it is synthesized in a laboratory, or when it
is tested in animals or humans? The term “developed” may
suggest a later stage in the multistep, highly regulated process
of bringing a drug to market, yet it likewise lacks a clear limiting
16
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
principle. Does development occur after early clinical studies,
or only after large-scale clinical trials are completed? The
absence of any clear limiting principle illustrates the difficulty
of imposing negligence liability untethered from traditional
defect-based principles.
The same uncertainty surrounds the Court of Appeal’s
requirement that the manufacturer “know[]” the alternative
drug is safer than and as effective as the existing drug. (Gilead
Tenofovir Cases, supra, 98 Cal.App.5th at p. 922.) This
requirement contemplates not only knowledge that the
alternative drug is safe and effective, but also comparative
knowledge that it is both safer and at least as effective as the
existing drug. The court accepted as true plaintiffs’ allegations
that Gilead possessed such knowledge based on a single phase
I/II clinical trial comparing TAF to TDF — a study lasting only
two weeks and involving just 30 subjects. But that conclusion
rests on a flawed premise; namely, that a drug can be
sufficiently developed, and its safety and efficacy sufficiently
known, before completion of phase III clinical trials and FDA
approval. That premise is inconsistent with the statutory and
regulatory framework governing drug development. It also
underscores the difficulty of imposing negligence liability based
on preliminary and evolving safety assessments of an
alternative drug still under development, rather than on a
defect in the drug currently being sold.
As the highly regulated drug development process makes
clear, drugs that appear promising in phase I or phase II trials
may later prove less effective than anticipated or may reveal
previously undetected adverse effects when studied over longer
periods and in larger, more diverse patient populations. (See
Step 3: Clinical Research, supra; 22 Case Studies, supra, at p. 2.)
17
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
The FDA cautions that early-stage trials may fail to detect less
common or long-term side effects, and that phase III trials are
more likely to uncover such risks because of their size and
duration. (Step 3: Clinical Research, supra.) As one amicus
curiae observes, positive early-phase results are better
understood as a basis for continued testing than as reliable
confirmation that a drug is safe, effective, or likely to obtain
regulatory approval. Moreover, even if an alternative drug
ultimately proves safer than an existing one for some patients,
it may present different or offsetting risks for others, further
undermining the premise that a manufacturer could reliably
“know” during early-stage development that an alternative drug
is categorically safer and equally effective for all users.2
Federal law reinforces this point. Under the FDCA, the
FDA may not approve a new drug absent “ ‘substantial
evidence’ ” of safety and efficacy, defined as “evidence consisting
of adequate and well-controlled investigations, including clinical
investigations.” (21 U.S.C. § 355(d).) The FDA has interpreted
this requirement to mean, in the ordinary case, “at least two
adequate and well-controlled studies” — typically phase III
clinical trials — demonstrating both safety and efficacy. (FDA,
Guidance for Industry: Providing Clinical Evidence of
Effectiveness for Human Drug and Biological Products (May
2
The dissent asserts that “plaintiffs’ claim is based on a
duty that exists in the latter stages of development.” (Dis. opn.
of Evans, J., post, at p. 9.) That description is difficult to
reconcile with the record. Plaintiffs contend Gilead acted
unreasonably after completing a two-week initial phase I/II
clinical trial involving only 30 subjects (only 20 of whom received
TAF), with approximately five additional years of clinical
testing, regulatory review, and commercialization efforts
remaining before TAF ultimately reached the market.
18
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
1998) p. 3 (hereafter Guidance for Industry)
<https://www.fda.gov/regulatory-information/search-fda-
guidance-documents/providing-clinical-evidence-effectiveness-
human-drug-and-biological-products> [as of August 3, 2026];
see also, 22 Case Studies, supra, at p. 2 [“Typically, a candidate
drug is submitted to the FDA for marketing approval after
phase 3 testing”].) Although the agency may, in limited
circumstances, rely on a single trial with confirmatory evidence,
that exception underscores the general rule rather than
displacing it. (See Guidance for Industry, at p. 3.) Consistent
with this framework, manufacturers are prohibited from
making claims about a new drug’s safety or efficacy prior to FDA
approval; such claims would be treated as false or misleading
statements. (21 U.S.C. § 352(n) & (gg)(1)(B)(i)(II); 21 C.F.R.
§ 202.1(e)(6)(i)–(ii), (xvi) (2022).) The statutory and regulatory
scheme thus reflects the understanding that definitive
judgments regarding a drug’s risks and benefits ordinarily
cannot be made before completion of phase III trials and FDA
approval. The Court of Appeal’s knowledge requirement is
therefore largely unattainable during the early stages of
development, as TAF was during the relevant period here. That
conclusion reinforces our doubt that liability may be imposed
under section 1714 for injuries caused by a concededly
nondefective drug based on allegedly unreasonable development
decisions regarding a different, unapproved drug.
The Court of Appeal believed that our decision in Mexicali
Rose v. Superior Court (1992) 1 Cal.4th 617 “effectively
resolve[d]” the issue by holding that a negligence claim may
proceed against a manufacturer absent a product defect.
(Gilead Tenofovir Cases, supra, 98 Cal.App.5th at p. 926.) In
Mexicali Rose, we held restaurants must “exercise reasonable
19
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
care in the preparation of [their] food” even where the food itself
is not defective. (Mexicali Rose, at p. 633.) But that holding
arose in the unique context of food preparation. Prepared food
may contain natural substances — there, a chicken bone — that
are inherent in food yet potentially dangerous when present in
a size, shape, or quantity capable of causing injury when
consumed. (Id. at p. 632.) Although it would make little sense
to characterize such natural substances as “defects” in the food,
it does make sense to require restaurants to exercise reasonable
care in removing them once they pose a sufficient danger.
Reflecting the narrow basis for our holding, we expressly limited
it “to commercial restaurant establishments.” (Id. at p. 619, fn.
1.) That context is far removed from prescription-drug
development. Thus, Mexicali Rose does not stand for the
broader principle that manufacturers may be liable in
negligence for injuries resulting from nondefective products.
For these reasons, we have substantial doubt that
California law recognizes a general negligence duty requiring
drug manufacturers to act reasonably in making drug-
development and commercialization decisions apart from their
established duty to design, manufacture, and market products
free from defects. Products liability law has long tied a
manufacturer’s duty of care to defects in the product placed into
the stream of commerce, and plaintiffs identify no clear limiting
principle for imposing liability absent proof of a product defect.
While we are skeptical that the duty plaintiffs propose is
consistent with products liability law as it has developed to this
point, we are cognizant that tort law is a broad and ever-
developing area of law. In this case, we need not definitively
decide whether the defect requirement categorically limits the
scope of a manufacturer’s general duty of reasonable care under
20
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
section 1714 in all circumstances. Even assuming a broader
duty could exist in some circumstances, plaintiffs’ theory — that
Gilead owed them a duty to make reasonable development and
commercialization decisions regarding TAF — is not cognizable
in light of the foreseeability and public policy considerations
identified in Rowland, supra, 69 Cal.2d at page 113. We turn to
the Rowland factors below.
C. Duty of care
In Rowland, we identified several considerations that may
justify a judicial exception to section 1714’s duty of care.
(Rowland, supra, 69 Cal.2d at pp. 112–113.) These
considerations fall into two categories. The foreseeability
factors evaluate “the foreseeability of harm to the plaintiff,” as
well as related concerns such as “the degree of certainty that the
plaintiff suffered injury” and “the closeness of the connection
between the defendant’s conduct and the injury suffered.” (Id.
at p. 113; accord, Kuciemba, supra, 14 Cal.5th at p. 1021.) The
public policy factors examine “the moral blame attached to the
defendant’s conduct, the policy of preventing future harm, the
extent of the burden to the defendant and consequences to the
community of imposing a duty to exercise care with resulting
liability for breach, and the availability, cost, and prevalence of
insurance for the risk involved.” (Rowland, at p. 113; accord,
Kuciemba, at pp. 1021–1022.)
Different timeframes apply to the two sets of factors.
(Kuciemba, supra, 14 Cal.5th at p. 1022.) The foreseeability
factors “are assessed based on information available during the
time of the alleged negligence,” whereas policy considerations —
such as the burdens imposed on defendants — are forward-
looking. (Ibid.) Both the foreseeability and the public policy
21
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
factors are analyzed “ ‘at a relatively broad level of factual
generality’ ” to determine not whether an exception is
warranted in the specific circumstances of a particular case, but
rather “ ‘whether carving out an entire category of cases from
that general duty rule is justified by clear considerations of
policy.’ ” (Id. at p. 1021.) In short, “ ‘the duty analysis is
categorical, not case specific.’ ” (Ibid.)
Nevertheless, while the Rowland factors are analyzed at a
categorical level rather than based on case-specific facts, we may
properly consider the general circumstances regarding the
development process for TAF — including the steps Gilead
needed to take to bring TAF to market — in determining
whether manufacturers in similar positions should owe a duty
of care to users of an existing drug when deciding whether and
when to commercialize an alternative drug. We do not, however,
consider Gilead’s particular internal deliberations, resource
constraints, or other case-specific factors that may have
influenced its decision regarding whether and when to
commercialize TAF. Such fact-bound considerations are not
part of the court’s duty analysis and are instead properly
reserved for the fact finder in determining whether a duty was
breached.
We analyze the Rowland factors to determine whether
foreseeability and public policy considerations warrant an
exception to any duty of reasonable care a drug manufacturer
may otherwise owe to users of a nondefective drug when
22
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
developing and deciding whether to commercialize a potentially
safer alternative.
1. Foreseeability Factors
Of the three foreseeability factors — foreseeability,
certainty, and the closeness of the connection between the
defendant’s conduct and the injury — the most important is the
first one; that is, “whether the injury in question was
foreseeable.” (Kesner v. Superior Court (2016) 1 Cal.5th 1132,
1145 (Kesner).) Our task “ ‘is not to decide whether a particular
plaintiff’s injury was reasonably foreseeable in light of a
particular defendant’s conduct, but rather to evaluate more
generally whether the category of negligent conduct at issue is
sufficiently likely to result in the kind of harm experienced that
liability may appropriately be imposed.’ ” (Cabral, supra,
51 Cal.4th at p. 772.) We therefore consider whether it was
reasonably foreseeable that a drug manufacturer’s decision to
delay taking the steps necessary to develop and commercialize
an allegedly safer drug would result in harm to users of the
existing drug. Because almost any harm can be described as
foreseeable after the fact, our analysis focuses on whether such
harm was reasonably foreseeable to the manufacturer at the
time it chose to delay developing and commercializing the new
drug. To do otherwise would risk hindsight bias, since “ ‘with
hindsight, everything is foreseeable.’ ” (Colonial Van & Storage,
Inc. v. Superior Court (2022) 76 Cal.App.5th 487, 503; accord,
Vasilenko v. Grace Family Church (2017) 3 Cal.5th 1077, 1088
(Vasilenko).)
The Court of Appeal concluded that this factor did not
support recognizing an exception to the duty of care based on
the premise that a drug manufacturer is capable of knowing,
23
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
before completing phase III clinical trials and obtaining FDA
approval, that an alternative drug is both safer than and as
effective as an existing drug. (See Gilead Tenofovir Cases,
supra, 98 Cal.App.5th at pp. 937, 945–948.) But as explained
above, the difficulty of making comparative judgments, the
structure of the drug-approval process, and the governing
statutory and regulatory framework demonstrate that such
knowledge ordinarily cannot be established at earlier stages of
development.
Early clinical studies are preliminary and provide only
limited, provisional information about a new drug’s safety and
effectiveness. Phase I trials involve small numbers of healthy
volunteers or patients with the targeted condition and are
designed primarily to assess basic safety, tolerability, and
dosage information, rather than overall therapeutic benefits
and risks. (21 C.F.R. § 312.21(a) (2022); Step 3: Clinical
Research, supra; 22 Case Studies, supra, at p. 2.) Phase II trials
expand testing to a somewhat larger group of patients with the
targeted condition, but they remain exploratory in nature and
focus primarily on short-term side effects rather than long-term
safety. (21 C.F.R. § 312.21(b) (2022); Step 3: Clinical Research,
supra; 22 Case Studies, supra, at p. 2.) The results of these
studies may appear promising, only for further research to
reveal previously unknown long-term or rare side effects that
may cause the new drug’s risks to outweigh its benefits. (Step
3: Clinical Research, supra; 22 Case Studies, supra, at p. 3.)
Indeed, according to the FDA, only about one-third of
drugs evaluated in phase II clinical trials yield results
promising enough to proceed to phase III testing. (Step 3:
Clinical Research, supra.) And only about 25 to 30 percent of
drugs that advance to phase III trials are considered successful
24
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
enough to undergo further testing or to present results to the
FDA for approval. (Ibid.) Drugs that appear promising in phase
I or phase II trials may later prove less effective than
anticipated or reveal previously undetected adverse effects
when studied for a longer period in larger and more diverse
patient populations. (See Step 3: Clinical Research, supra;
22 Case Studies, supra, at p. 2.) Consistent with this
understanding, the FDCA requires “substantial evidence” of
safety and efficacy, generally from at least two adequate and
well-controlled studies — typically phase III clinical trials —
before a drug may receive FDA approval. (21 U.S.C. § 355(d);
Guidance for Industry, supra, at p. 3.) Manufacturers may not
represent an investigational drug as safe or effective before FDA
approval. (21 U.S.C. § 352(n) & (gg)(1)(B)(i)(II); 21 C.F.R.
§ 202.1(e)(6)(i)–(ii), (xvi) (2022).)
The dissent adopts an approach that effectively removes
this inquiry from the court’s purview by asserting that the
knowledge question must be reserved for the fact finder. (Dis.
opn. of Evans, J., post, at p. 15.) But Rowland requires courts
to determine, as a matter of law, whether drug manufacturers
as a class could reasonably know, based on only early-phase
clinical trials, that an alternative drug is safer than an existing
one such that pausing the alternative drug’s development would
foreseeably harm users of the existing drug. The very purpose
of Rowland, which our courts have now applied for almost 60
years, is to allow courts — not a jury — to make a legal
determination as to whether liability can attach at all. If the
relevant knowledge inquiry were invariably reserved for the fact
finder, courts could never meaningfully assess foreseeability
25
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
based on “information available during the time of the alleged
negligence.” (Kuciemba, supra, 14 Cal.5th at p. 1022.)
For these reasons, the Court of Appeal erred in concluding
it is foreseeable that delaying the development and
commercialization of a drug still in the early phases of clinical
testing will harm users of an existing drug. (Gilead Tenofovir
Cases, supra, 98 Cal.App.5th at p. 938.) Because a
manufacturer cannot truly know of a drug’s safety and efficacy
before phase III testing and FDA approval, it cannot reasonably
foresee early in the development process that any delay in
commercializing an alternative drug will harm users of an
existing one. This factor therefore weighs in favor of recognizing
an exception to any duty of care manufacturers might otherwise
owe when developing and commercializing potentially safer
alternative drugs.
The second foreseeability factor is “the degree of certainty
that the plaintiff suffered injury.” (Rowland, supra, 69 Cal.2d
at p. 113.) This factor applies “ ‘primarily, if not exclusively,
when the only claimed injury is an intangible harm, such as
emotional distress,’ ” because such harms can be uncertain or
difficult to verify. (Kuciemba, supra, 14 Cal.5th at p. 1023.)
Here, plaintiffs’ personal injury claims “are both tangible and
amenable to compensation.” (Ibid.) Because plaintiffs allege
concrete physical injuries that are readily compensable, this
factor does not support recognizing an exception to a duty of
care.
The third foreseeability factor requires us to examine “the
closeness of the connection between the defendant’s conduct and
the injury suffered.” (Rowland, supra, 69 Cal.2d at p. 113.)
“[W]here the injury suffered is connected only distantly and
26
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
indirectly to the defendant’s negligent act, the risk of that type
of injury from the category of negligent conduct at issue is likely
to be deemed unforeseeable.” (Cabral, supra, 51 Cal.4th at
p. 779.) This factor therefore requires an examination of the
causal connections between the alleged negligence and the
injury, an inquiry “akin to analyzing proximate causation.”
(Kuciemba, supra, 14 Cal.5th at p. 1024.) In addition, it
“accounts for third party or other intervening conduct” that is
independent of, and not derivative of, the defendant’s
negligence. (Vasilenko, supra, 3 Cal.5th at p. 1086.)
A drug manufacturer’s decision to delay the
commercialization of a new drug is, at most, only remotely
connected to a plaintiff’s injury allegedly caused by use of an
existing drug. (See, e.g., Vasilenko, supra, 3 Cal.5th at p. 1086
[a church’s directions to park across a public street bore only an
attenuated relationship to an injury resulting from the
independent decisions by the plaintiff and a third party driver];
State Dept. of State Hospitals v. Superior Court (2015)
61 Cal.4th 339, 355–356 (State Hospitals) [no proximate
causation where the injury depended on a speculative chain of
discretionary third party decisions, including a third party
evaluator’s determination as to whether the injury-causing
actor was a sexually violent predator].) Any such injury would
depend on speculation regarding the independent conduct and
discretionary decisions of multiple third party actors, as well as
the uncertain outcomes of additional clinical trials and scientific
research required to develop, approve, and market the new
drug.
First, as explained above, early clinical studies provide
only preliminary and incomplete information regarding a drug’s
safety and efficacy, and drugs that initially appear promising
27
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
may later prove ineffective or reveal previously undetected
adverse effects during phase III trials involving larger and more
diverse patient populations. (See Step 3: Clinical Research,
supra; 22 Case Studies, supra, at p. 2.) Thus, any causal
connection between a drug manufacturer’s decision to pause or
discontinue development of an alternative drug after having
completed only early-phase clinical testing and a plaintiff’s later
injury from an existing drug may be severed if the alternative
drug ultimately fails in phase III trials.
Second, if the manufacturer decides to proceed to phase III
testing and the results of such testing are promising, the drug
may nevertheless fail to obtain FDA approval. (See Sacks et al.,
Scientific and Regulatory Reasons for Delay and Denial of FDA
Approval of Initial Applications for New Drugs, 2000-2012 (Jan.
2014) 311 JAMA 4 (hereafter FDA Approval) [analyzing cases in
which the FDA declined to approve a new drug because it failed
to demonstrate superior efficacy or increased safety]
<https://jamanetwork.com/journals/jama/fullarticle/1817795>
[as of August 3, 2026].) Plaintiffs do not dispute Gilead’s
assertion that, “of medicines entering clinical trials, fewer than
one out of eight will obtain FDA approval.” (Gilead Tenofovir
Cases, supra, 98 Cal.App.5th at p. 939; see also Step 3: Clinical
Research, supra; United States Government Accountability
Office, Investigational New Drugs: FDA Has Taken Steps to
Improve the Expanded Access Program but Should Further
Clarify How Adverse Events Data Are Used (July 2017) p. 6 [of
the drugs that entered phase I trials, 9.6 percent received FDA
approval] <https://www.gao.gov/assets/gao-17-564.pdf> [as of
August 3, 2026].)
The Court of Appeal did not find this fact significant,
opining that because the asserted duty involves a drug
28
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
candidate “the manufacturer knows to be as effective as, and
safer than, an existing drug” “FDA approval [is] far less
uncertain than might otherwise be the case.” (Gilead Tenofovir
Cases, 98 Cal.App.5th at p. 939.) But again, the court erred in
accepting that a drug manufacturer can genuinely know of a
new drug’s safety and efficacy before completing phase III
clinical trials. Moreover, even where a manufacturer goes on to
complete phase III trials and, based on the results of those
studies, believes that its new drug is promising and likely to
obtain FDA approval, the FDA may nevertheless disagree and
deny approval. (See FDA Approval, supra, 311 JAMA 4.) Thus,
a manufacturer’s own assessment of a drug’s safety and efficacy
does not make FDA approval certain. (Ibid.)
Third, if the FDA approves the new drug and it is brought
to market, a patient’s physician may nevertheless decide, in
consultation with the patient, that the existing drug is the better
option for that individual. The Court of Appeal discounted this
possibility, reasoning that “once the FDA has approved an
alternative that is safer and at least equally effective for the
patient concerned, the manufacturer would reasonably expect
doctors to prescribe the new medication in place of the old.”
(Gilead Tenofovir Cases, supra, 98 Cal.App.5th at p. 939.) That
conclusion, however, rests on the assumption that a new drug is
a safer and better option than an existing one for all patients —
an assumption that fails to account for the complexity of
prescription drug therapy. Typically, alternative drugs used to
treat the same condition present different risks and benefits,
making one drug safer for some patients but not necessarily for
others. We recognized this reality in Brown, where we observed
that “the superiority of one drug over another” must be
evaluated with respect to each individual plaintiff, “since the
29
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
advantages of a drug cannot be isolated from the condition of a
particular patient.” (Brown, supra, 44 Cal.3d at p. 1068.)
“Thus, in one case the drug that injured the plaintiff might be
the better choice, while this would not be true as to another
user.” (Ibid.) Physicians and patients may therefore reasonably
differ in their assessments of which drug represents the safer
option. For example, a new drug may carry a lower risk of a
severe side effect — such as liver failure — while also causing
other, less serious but more common side effects, such as weight
gain or increased cholesterol. A physician treating a patient
concerned about weight gain or high cholesterol might conclude
that the existing drug is the better choice, while another
physician might determine that the less severe side effects are
acceptable risks when weighed against the risk of liver failure.
Indeed, this variability in physician and patient
preferences is evident even with respect to TDF and TAF. As
plaintiffs conceded in their briefing to the Court of Appeal below,
“for a variety of reasons, some physicians and patients prefer
TDF over TAF.” And while plaintiffs may now believe that they
would not have opted to take TDF over TAF given the injuries
they allege they ultimately suffered, such retrospective
judgments are “prone to hindsight bias.” (Himes v. Somatics,
LLC (2024) 16 Cal.5th 209, 234 [rejecting subjective patient
standard for failure-to-warn claims against prescription drug
manufacturers due to risk of hindsight bias]; accord, Cobbs v.
Grant (1972) 8 Cal.3d 229, 245 [rejecting subjective patient
standard for informed consent claims due to risk of hindsight
bias].)
The dissent fails to engage with this inquiry. Instead, it
mischaracterizes our analysis as suggesting that independent
decisions by the FDA, physicians, or others “worsened” or
30
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
“exacerbated” plaintiffs’ injuries (dis. opn. of Evans, J., post, at
p. 16), when our point is that those intervening decisions render
the connection between the alleged conduct and plaintiffs’
injuries too remote to make the risk of injury sufficiently
foreseeable. (See Cabral, supra, 51 Cal.4th at p. 779 [“[W]here
the injury suffered is connected only distantly and indirectly to
the defendant’s negligent act, the risk of that type of injury from
the category of negligent conduct at issue is likely to be deemed
unforeseeable”].) If the alternative drug failed in phase III
trials, or was not approved by the FDA, or would not have been
prescribed to a plaintiff, then there would be no connection
between a drug manufacturer’s decision to pause development
of that drug and any injuries resulting from the use of an
existing drug.
For these reasons, and consistent with our prior decisions
applying the third foreseeability factor, we conclude that a drug
manufacturer’s decision to cease taking the steps necessary to
develop and bring to market an allegedly safer drug is remote
from a plaintiff’s injuries caused by an existing drug. (See
Cabral, supra, 51 Cal.4th at pp. 779–780; Vasilenko, supra,
3 Cal.5th at p. 1086.) Any asserted causal connection depends
on a speculative chain of intervening events and discretionary
decisions by third parties that are neither foreseeable in any
concrete sense nor derivative of the manufacturer’s conduct.
(See Vasilenko, at p. 1086; State Hospitals, supra, 61 Cal.4th at
pp. 355–356.) Those intervening acts include the uncertain
outcomes of additional clinical testing, the FDA’s independent
regulatory judgment whether to approve the new drug, and
physicians’ individualized prescribing decisions based on
patient-specific risk-benefit assessments. These independent
and contingent decisions render the connection between the
31
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
defendant’s conduct and the plaintiff’s injury highly attenuated.
Accordingly, this factor supports recognizing an exception to any
duty a manufacturer might otherwise owe to develop and
commercialize an allegedly safer drug.
2. Public Policy Factors
In addition to foreseeability factors, the Rowland analysis
includes policy considerations which “may dictate a cause of
action should not be sanctioned no matter how foreseeable the
risk” (Elden v. Sheldon (1988) 46 Cal.3d 267, 274), such as
where the consequences of recognizing a particular duty of care
would create “an intolerable burden on society” (id. at p. 277;
accord, Kuciemba, supra, 14 Cal.5th at p. 1025).
The first policy factor is “the moral blame attached to the
defendant’s conduct.” (Rowland, supra, 69 Cal.2d at p. 113.)
Moral blameworthiness is difficult to evaluate categorically in
this context because drug manufacturers, as a class, may delay
developing and commercializing an alternative drug for a
variety of reasons. The Court of Appeal correctly observed that
“a manufacturer’s decision to delay the commercialization of a
safer drug may be made for morally neutral, or even worthy,
reasons.” (Gilead Tenofovir Cases, supra, 98 Cal.App.5th at
p. 942.) As amicus curiae Community Education Group
explains, a manufacturer might prioritize allocating its
resources to developing a treatment for a disease or condition
for which no therapies currently exist, rather than developing a
safer therapy for a disease or condition already adequately
treated by a nondefective drug. Alternatively, a manufacturer
might determine that, although an alternative drug presents a
lower risk of certain rare but severe side effects, patients would
not prefer it because it requires a complex dosing schedule or
32
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
carries a higher risk of more common side effects. Or a
manufacturer might conclude that any safety improvements are
only marginal and that devoting resources to developing and
commercializing the new drug would not be worthwhile,
particularly where the existing drug was already meeting
patient needs. Manufacturers’ decisions about whether to
continue developing or commercializing an alternative drug
ordinarily involve competing scientific, medical, regulatory, and
business considerations, including resource allocation and
commercial viability. Especially in the early phases of drug
development, those decisions must be made with limited
information about which drug candidates are most likely to
succeed.
The Court of Appeal nevertheless concluded — somewhat
circuitously — that the moral blame factor favors recognizing a
duty of care because the asserted duty attaches only to a drug
manufacturer’s negligence in its decisions about
commercializing an allegedly safer drug. (Gilead Tenofovir
Cases, supra, 98 Cal.App.5th at p. 942.) But if negligence alone
sufficed to establish moral blameworthiness, this factor would
invariably favor the imposition of a duty of care in every case,
since negligence is alleged in every negligence action. For this
reason, “the moral blame that attends ordinary negligence is
generally not sufficient to tip the balance of the Rowland factors
in favor of liability.” (Adams v. City of Fremont (1998)
68 Cal.App.4th 243, 270, disapproved of on another ground in
USA Taekwondo, supra, 11 Cal.5th 204.) Indeed, we have found
little to no moral blameworthiness when a driver had
negligently stopped a vehicle alongside a freeway for a
nonemergency purpose (Cabral, supra, 51 Cal.4th at p. 782) and
when a church negligently required invitees to park in a lot
33
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
across a busy street from the church (Vasilenko, supra, 3 Cal.5th
at p. 1091).
Moral blameworthiness is instead found where the
defendant has acted not merely unreasonably but with some
degree of moral culpability, such as where it has actual
knowledge of a known but hidden danger and fails to take
reasonable measures to avert foreseeable harm (see Peterson v.
San Francisco Community College Dist. (1984) 36 Cal.3d 799,
814); reaps a financial benefit from risks it created (Kuciemba,
supra, 14 Cal.5th at p. 1025); exploits a relative inequality
between the parties or exercises greater control over the risks at
issue (id. at p. 1026); or engages in affirmative misconduct that
society seeks to deter (see Christensen v. Superior Court (1991)
54 Cal.3d 868, 896–898).
The dissent focuses on plaintiffs’ allegations that Gilead
deliberately delayed developing and commercializing TAF solely
to extend its patent over tenofovir-related medications and
maximize its profits. (Dis. opn. of Evans, J., post, at pp. 17–19.)
But Rowland requires us to evaluate moral blame categorically
rather than by reference to the precise circumstances and
motivations plaintiffs attribute to Gilead in this case. (Gilead
Tenofovir Cases, supra, 98 Cal.App.5th at p. 942; see also id. at
p. 936.) The question is not whether Gilead’s alleged reasons for
delay were proper or improper, but rather whether the category
of conduct alleged ordinarily reflects moral culpability. And
although, at a categorical level, profit considerations may be
among the many factors drug manufacturers weigh when
deciding whether to continue developing and commercializing
34
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
an alternative drug, such considerations may be only one among
several that guide the decision to delay development.
Citing to our holdings in Kesner and Kuciemba, the dissent
additionally asserts that moral blameworthiness must be found
whenever there exists “the reaping of a profit and inherent
power imbalance between plaintiffs and defendants.” (Dis. opn.
of Evans, J., post, at p. 18.) But that description applies to
virtually every negligence action brought against a business.
We do not read Kesner and Kuciemba so expansively. Although
in certain circumstances profit maximization can be
characterized as “immoral” in some sense, a manufacturer’s
decision to delay developing an allegedly safer version of a
lifesaving drug solely to increase profits is morally blameworthy
only if the manufacturer can reliably know, before later-stage
clinical testing and FDA approval, that the alternative drug is
in fact safer than the existing one. As explained above, it
cannot.
Finally, assuming a drug manufacturer could know —
prior to phase III testing and FDA approval — that an
alternative drug is safer than an existing one, and further
assuming manufacturers would categorically decide to delay
commercializing alternative drugs only to maximize profits, we
have explained that moral blameworthiness depends upon
whether “there were reasonable ameliorative steps the
defendant could have taken . . . ‘to avert the foreseeable harm.’ ”
(Vasilenko, supra, 3 Cal.5th at p. 1091.) For example, in
Parsons, supra, 15 Cal.4th 456, we concluded that a garbage
truck operator had no reasonable means of averting the risk that
a horse would be startled by its ordinary operations, because any
proposed precautions — including altering collection hours or
temporarily blocking the collection area — would be impractical
35
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
and socially costly. (Id. at p. 474.) Likewise, in Vasilenko, we
found no moral blame where it was unclear what effective and
affordable ameliorative steps a church could take to mitigate the
danger posed by requiring invitees to cross a busy street
between a parking lot and the church. (Vasilenko, at p. 1091.)
We noted that, while a landowner could petition relevant
authorities to install traffic control devices, such as crosswalks
or traffic signals, the ultimate decision rests with those
authorities. (Id. at p. 1087.)
Similarly, here, the ameliorative steps a drug
manufacturer would need to take to develop and bring an
alternative drug to market are both resource-intensive and
time-consuming, since extensive and costly clinical trials are
required to secure FDA approval. Moreover, because the FDA
is the ultimate decision maker as to whether a new drug may be
marketed, the timing and feasibility of commercializing an
alternative drug are not entirely within the manufacturer’s
control. For these reasons, we conclude that a drug
manufacturer’s negligent decision to delay developing and
commercializing an alternative drug is not categorically morally
blameworthy.
The second policy factor is “the policy of preventing future
harm.” (Rowland, supra, 69 Cal.2d at p. 113.) This policy “is
ordinarily served . . . by imposing the costs of negligent conduct
upon those responsible.” (Cabral, supra, 51 Cal.4th at p. 781.)
It may be outweighed, however, “by laws or mores indicating
approval of the conduct or by the undesirable consequences of
allowing potential liability.” (Id. at p. 782.) Accordingly, this
factor “examines both the positive and the negative societal
36
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
consequences of recognizing a tort duty.” (Kuciemba, supra,
14 Cal.5th at p. 1026.)
On the one hand, a duty of care in this context could
incentivize manufacturers to more promptly develop and
commercialize drugs that may prove to be safer than and as
effective as existing treatments, thereby preventing harm to
some users of those treatments. It is plainly in the public
interest for manufacturers to develop and bring to market safer
drugs that may reduce injuries to patients already suffering
from serious underlying conditions. Here, plaintiffs allege that
had they been able to take TAF rather than TDF at an earlier
date, they may have avoided serious side effects caused by TDF.
On the other hand, as plaintiffs themselves acknowledge,
drug manufacturers already have “ample incentive[s] to release
safer products into the marketplace,” including the opportunity
to gain market share and avoid reputational harm associated
with adverse effects. This existing incentive structure lessens
the need for tort liability as a means to ensure reasonable
decisionmaking about whether and when to develop safer
alternative drugs. (See Vasilenko, supra, 3 Cal.5th at p. 1088
[observing that landowners already have incentives to provide
safe and convenient parking].)
Perhaps more importantly, such a duty risks creating
perverse incentives that may undermine, rather than advance,
the goal of preventing future harm. Decisions regarding
whether to continue developing a promising drug, how quickly
to advance it through clinical trials, and how to allocate finite
research and development resources are inherently complex.
They rest on imperfect and evolving information, and often
involve competing scientific, regulatory, and public-health
37
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
considerations. As amicus curiae Community Education Group
explains, a duty to develop and commercialize, without undue
delay, a potentially safer alternative to a nondefective drug
could distort those decisions by encouraging manufacturers to
prioritize incremental improvements to existing therapies at the
expense of pursuing novel treatments for diseases currently
lacking effective treatments. Because the proposed duty would
attach only to alternative drugs that are purportedly safer than
existing therapies — and not to entirely new therapies — a
manufacturer would face potential tort liability only if it failed
to develop an alternative drug. A rational manufacturer could
therefore conclude that investing additional resources in
developing alternative drugs is necessary to reduce litigation
risk, while research directed toward entirely new therapies
carries no corresponding tort incentive.
Amicus curiae further observes that manufacturers may
become reluctant to initiate early clinical studies of backup or
alternative drug candidates for fear that those studies could
later be used to support negligence claims. Faced with the
possibility that a decision to continue developing a lead
candidate rather than a backup candidate could later be deemed
unreasonable, manufacturers might forgo early-stage research
of backup candidates altogether. Alternatively, if
manufacturers deem it too risky to wholly forgo researching
backup candidates — given the possibility that a lead candidate
may fail — they might instead delay commercializing a lead
candidate until all potential backup candidates are fully tested
in clinical trials.
This Court has previously declined to impose liability
rules that would produce precisely these outcomes. In Brown,
we rejected strict liability for design defects in prescription
38
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
drugs, cautioning that “[i]f drug manufacturers were subject to
strict liability, they might be reluctant to undertake research
programs to develop some pharmaceuticals that would prove
beneficial or to distribute others that are available to be
marketed,” and might instead withhold a drug from the market
“until scientific skill and knowledge advanced to the point at
which additional dangerous side effects would be revealed.”
(Brown, supra, 44 Cal.3d at p. 1063.) Similarly, a duty not to
unreasonably delay commercializing alternative treatments
could distort manufacturers’ decisions regarding whether to
continue developing such treatments where preliminary
research suggests they may reduce certain risks, but the
manufacturers would otherwise elect not to pursue them due to
competing concerns or research priorities.
Plaintiffs argue that these concerns are overstated
because liability would attach only to unreasonable decisions
regarding whether and when to commercialize an allegedly
safer drug. They contend, for example, that allocating resources
to address a more pressing medical need would not be
considered unreasonable. That may be true in the abstract, but
it would likely prove difficult to evaluate the reasonableness of
a drug manufacturer’s resource allocation decisions in
hindsight. Amicus curiae Pharmaceutical Research and
Manufacturers of America explains that drug development
frequently involves pursuing multiple potential drug candidates
in parallel, with the understanding that many drug candidates
will fail. As noted above, plaintiffs do not dispute Gilead’s
contention that only one out of eight drug candidates entering
clinical trials ultimately receives FDA approval. (Gilead
Tenofovir Cases, supra, 98 Cal.App.5th at p. 939.) According to
amicus curiae, because of this high failure rate, it is standard
39
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
industry practice to collect data on backup candidates while
proceeding with a lead candidate. During this development
process, drug manufacturers must make complex strategic
decisions about where to devote finite resources based on limited
information. These decisions are often provisional, iterative,
and contingent on preliminary laboratory or clinical studies that
may later prove incorrect. As a result, they do not readily lend
themselves to retrospective evaluations of reasonableness
armed with the benefit of hindsight.
To illustrate, a manufacturer might determine, based on
early clinical data, that a backup candidate had the potential to
reduce certain side effects associated with a lead candidate. The
manufacturer may nevertheless discontinue development if it
believes patients would prefer the lead drug because of dosing
convenience or due to other side effects associated with the
backup drug. If the evidence later reveals that those concerns
were unfounded and that patients ultimately fared better on the
backup drug, a retrospective negligence inquiry could conclude
that the manufacturer’s initial development decision was
unreasonable. When multiple potential alternative drugs exist,
moreover, there is no obvious basis for determining whether one
development path was more reasonable than another.
As explained above, assessing the reasonableness of a
drug manufacturer’s research and development decisions differs
fundamentally from evaluating product defect claims. Product
defect claims are judged based on contemporaneous and
relatively complete scientific evidence available at the time of
manufacture and distribution. (See Carlin, supra, 13 Cal.4th at
p. 1116.) By contrast, evaluating drug development choices
entails a backward-looking assessment of the research priorities
a manufacturer adopted based on preliminary and evolving
40
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
scientific evidence. This retrospective inquiry creates a
substantial risk that hindsight will supplant the manufacturer’s
contemporaneous judgment, even when its decisions were
reasonable when made.
Contrary to the dissent’s suggestion, we do not question a
fact finder’s ability to resolve ordinary negligence claims
involving complex scientific evidence, such as negligent design,
manufacture, or marketing claims. (Dis. opn. of Evans, J., post,
at pp. 10–12.) Rather, the Rowland inquiry requires us to
determine whether this category of conduct should be subjected
to negligence liability in the first place. Unlike traditional
products liability claims, this novel theory would require fact
finders to assess whether a manufacturer acted reasonably in
deciding whether and when to continue researching, developing,
and commercializing a drug that never completed clinical
testing or received FDA approval. That inquiry asks fact finders
to reconstruct years of research and development decisions
made in the face of evolving scientific evidence and uncertain
outcomes, rather than to evaluate the safety of a product that
actually reached the market. Neither the plaintiffs nor the
dissent have identified a single case wherein a fact finder has
been tasked with making such determinations.
In sum, although recognizing a tort duty might reduce
injuries for some patients, it also carries a risk of adverse
consequences that may not serve public health and safety as a
whole. We accordingly conclude that this factor weighs in favor
of recognizing an exception to any duty of care a drug
manufacturer may otherwise owe in this context.
The third policy factor examines “the extent of the burden
to the defendant and consequences to the community of
41
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
imposing a duty to exercise care with resulting liability for
breach.” (Rowland, supra, 69 Cal.2d at p. 113.) This factor
overlaps somewhat with the prior one, insofar as it requires
consideration of potential adverse consequences to the
community. (Lorenzo v. Calex Engineering, Inc. (2025)
110 Cal.App.5th 49, 70; see also Vasilenko, supra, 3 Cal.5th at
p. 1090.)
The asserted duty may undermine, rather than promote,
public health and safety by inviting retrospective second-
guessing of complex development decisions made under
conditions of scientific uncertainty. As Gilead and several amici
curiae explain, a manufacturer may develop a highly effective
and lifesaving drug and, in the ordinary course of research and
development, conduct preliminary studies of backup or
alternative candidates — both to pursue potential
improvements and to hedge against the substantial risk that
any single candidate will fail. If early data on a backup
candidate suggests a marginal reduction in certain severe side
effects for a limited subset of patients, the asserted duty might
expose the manufacturer to significant liability for failing to
commercialize that alternative, notwithstanding its limited
benefits or offsetting risks. This may create perverse incentives:
manufacturers might decline to investigate backup or
alternative candidates at all, lest preliminary indications of
incremental safety improvements later be used to challenge the
reasonableness of their development decisions. That result is
especially troubling in the pharmaceutical context where, as
Gilead and several amici curiae explain, high clinical-failure
rates make parallel development of backup candidates standard
practice and essential to avoiding delays in the availability of
beneficial medicines if a lead candidate fails. Alternatively,
42
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
because liability would only attach once a manufacturer
releases a drug, manufacturers may be incentivized to delay
commercialization of an otherwise nondefective and potentially
lifesaving drug until all backup candidates have been fully
explored, slowing patient access while companies search for a
hypothetical “perfect” drug.
The asserted duty would also risk interfering with
pharmaceutical innovation and the allocation of finite research
and development resources more broadly. As Gilead and amicus
curiae Community Education Group observe, the duty may
encourage manufacturers to focus on making marginal
improvements to existing drugs rather than pursuing novel
therapies for diseases that lack effective treatment, or to rush to
market alternatives that appear safer for some patients but pose
greater risks for others. Indeed, the asserted duty may force
manufacturers to make an untenable choice: either devote
scarce research and development resources to advancing a
backup drug to an existing treatment that may reduce the risk
of rare side effects for a subset of patients, or risk liability by
prioritizing developing a new drug to treat a currently
untreatable disease that may prove to be lifesaving for a larger
population. More generally, as amicus curiae International
Association of Defense Counsel points out, manufacturers could
be driven to scour their development histories to identify all
conceivable alternative drugs that may reduce risk for any user
under any foreseeable circumstance of use, no matter how
remote. To avoid liability, manufacturers may then feel
pressured to divert limited resources toward commercializing
those alternatives, regardless of their overall risk-benefit
43
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
profiles or the substantial costs of development and
commercialization.
The Court of Appeal rejected these concerns, dismissing
Gilead’s argument that a duty of care in this context would place
manufacturers “ ‘under an endless obligation to pursue ever-
better new products or improvements to existing products.’ ”
(Gilead Tenofovir Cases, supra, 98 Cal.App.5th at p. 921.) The
court reasoned that the duty merely requires manufacturers to
act reasonably in deciding whether to commercialize a drug they
have already invented and know to be safer than an existing
one. (See id. at pp. 921–922, 944.) In reality, however, the duty
would likely pressure manufacturers to promptly commercialize
all alternative drugs that could later be characterized as safer,
even when the existing product is concededly nondefective.
Whenever preliminary research suggests that an alternative
might reduce certain side effects, a fact finder could be asked to
make a backwards-looking determination as to whether the
manufacturer acted unreasonably in failing to bring that
alternative to market sooner. For the reasons explained above,
this duty threatens to skew innovation priorities in a manner
that does not advance — and may ultimately undermine —
public health and safety.
Moreover, plaintiffs provide no legal basis for confining
their theory of liability to the pharmaceutical industry. If
manufacturers across all industries owe a general duty of care
in making decisions about whether and when to develop and
commercialize allegedly safer products, then manufacturers of a
wide range of complex consumer products could face liability for
their development and commercialization decisions even when
the product at issue is otherwise reasonably safe and
nondefective. For example, an automobile manufacturer that
44
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
does not install particular safety technology — such as lane-
change assist or automatic emergency braking — on its vehicles
could be exposed to liability for not making the technology
available sooner, notwithstanding the fact that its vehicles meet
all applicable safety standards and are deemed nondefective
under the risk-benefit test. Manufacturers might avoid
researching or testing potential safety improvements altogether
for fear that preliminary findings might later be used to allege
negligence. While we do not decide here whether manufacturers
outside of the pharmaceutical context have a duty to
commercialize allegedly safer products to replace concededly
non-defective ones, we note that accepting plaintiffs’ theory of
liability could have far-reaching consequences across a broad
range of industries.
Not only would a duty of care risk undermining public
safety, it would also impose substantial burdens on drug
manufacturers. In analyzing this factor, the Court of Appeal
focused primarily on Gilead’s argument that a duty would result
in a “flood of lawsuits.” (Gilead Tenofovir Cases, supra,
98 Cal.App.5th at p. 944.) Although the court concluded that
Gilead “overstate[d]” this concern (ibid.), it acknowledged that
the duty could extend to “a potentially large class of persons” (id.
at p. 945). The Court of Appeal nevertheless concluded —
without further explanation — that “most plaintiffs would likely
face a difficult road in establishing a breach of the duty of
reasonable care.” (Id. at p. 944.) It is far from clear, however,
that manufacturers’ decisions in this context would routinely be
found to be reasonable and nonnegligent, given that such
determinations hinge on inherently uncertain judgments, as
described above, about when an alternative drug is sufficiently
developed, whether later-stage clinical trials are likely to be
45
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
successful, how much delay in commercializing the alternative
drug is tolerable, and how to prioritize competing therapies.
The dissent contends that this duty would impose little
burden because manufacturers have “theoretically” always been
subject to it. (Dis. opn. of Evans, J., post, at p. 28.) But that
assertion conflates section 1714’s general duty of care with the
unprecedented theory of liability plaintiffs seek to impose here.
Neither the dissent nor plaintiffs identify a single decision
recognizing a duty requiring manufacturers to develop and
commercialize without delay an alternative drug to replace a
concededly nondefective one. This bears repeating: not once in
the dissent’s 29 pages does it cite a single case in this state or
any other finding liability on such a theory. The closest it comes
is our decision in Mexicali Rose — a case involving a
restaurant’s negligent failure to remove a harmful chicken bone
from an enchilada. (Dis. opn. of Evans, J., post, at pp. 6–7.) But,
for the reasons described above, this case bears little
resemblance to plaintiffs’ theory of liability. Moreover, it makes
little sense to premise an entirely new form of liability in a
dispute concerning complex life-saving drugs on a case involving
a chicken bone in an enchilada.
The dissent’s remaining out-of-state authorities likewise
miss the mark. They involve traditional negligent design,
manufacturing, marketing, or misrepresentation claims
challenging allegedly dangerous products — claims our decision
leaves untouched. (See dis. opn. of Evans, J., post, at pp. 29, 31;
see also State v. Purdue Pharma L.P. (R.I., Aug. 16, 2019, No.
PC-2018-4555) 2019 WL 3991963, p. *15 [recognizing duty “ ‘to
exercise reasonable care in manufacturing, marketing, selling,
and distributing highly dangerous opioid drugs’ ”]; Ohio County
Commission v. Express Scripts, Inc. (N.D.W. Va., Dec. 23, 2024,
46
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
No. 5:24-CV-142) 2024 WL 5701504, p. *15 [alleging defendants
collaborated with drug manufacturers to grossly misrepresent
addictive qualities of opioids]; Lance v. Wyeth (2014) 624 Pa.
231, 277 [drug manufacturer marketed drug it allegedly knew
was “too harmful to be used by anyone”].)
The dissent assumes that recognizing plaintiffs’ proposed
duty “would not disincentivize the development of drugs but
merely allow for the possibility of liability when the
development of less harmful alternatives is unreasonably
delayed in order to maximize profit.” (Dis. opn. of Evans, J.,
post, at p. 23.) In other words, the dissent appears to conclude
that a holding from this court which, for the first time, subjects
a drug manufacturer’s research and development decisions
regarding an alternative drug to potentially enormous tort
liability would have little effect on manufacturers’ willingness
to pursue alternative forms of treatment. We disagree that
manufacturers’ incentives would be unaffected by the threat of
such liability.
Moreover, we have said that “the most relevant burden is
the cost to the defendants of upholding, not violating, the duty
of ordinary care.” (Kesner, supra, 1 Cal.5th at p. 1152.) The
duty plaintiffs propose is particularly burdensome because it is
not merely a duty to market an already developed and approved
safer drug. Instead, it is a duty to continue researching and
developing an alternative drug whenever preliminary data
suggests it might ultimately prove safer than an existing one.
According to the dissent, Gilead may be liable simply because it
“paused development” after receiving supposedly “promising”
results from a two-week clinical trial involving only 30 subjects.
(Dis. opn. of Evans, J., post, at p. 9.) Under the threat of such
liability, manufacturers will be compelled to invest substantial
47
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
sums and devote significant resources to developing backup or
alternative drugs whenever preliminary research suggests a
potential reduction in certain adverse side effects associated
with an existing drug — or else risk a fact finder later
concluding that their decision not to do so was unreasonable.
This would require manufacturers to undertake large-
scale clinical testing. Plaintiffs’ and the dissent’s theory is that
Gilead’s 30-person Phase I/II study effectively required it to
conduct a full-scale Phase III trial. Unlike phase I or phase II
clinical trials, which typically involve fewer subjects studied
over a more limited timeframe, phase III clinical trials are far
more expansive. (See Step 3: Clinical Research, supra.) They
often enroll thousands of participants, can take several years to
complete, and may cost millions of dollars. (See ibid.; Moore et
al., Estimated Costs of Pivotal Trials for Novel Therapeutic
Agents Approved by the US Food and Drug Administration,
2015-2016 (Nov. 2018) 178 JAMA Internal Medicine 11 [finding
median costs of phase III clinical trials for new therapeutic
agents to be $19 million] <https://jamanetwork.com/journals/
jamainternalmedicine/fullarticle/2702287> [as of August 3,
2026].) Even if the results of phase III trials look promising,
success is far from guaranteed: The FDA estimates only 25 to
30 percent of drugs tested in phase III trials will move on to the
next development stage of seeking FDA approval. (Step 3:
Clinical Research, supra.) And manufacturers must then incur
additional costs to obtain FDA approval and bring the
alternative drug to market. As a result, the duty would require
manufacturers to invest years of time and large sums of money
to develop drugs that may ultimately yield no benefit to anyone.
In this case, for example, Gilead estimated projected costs
of approximately $100 million to conduct additional clinical
48
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
trials on TAF and complete the steps necessary to obtain FDA
approval and bring the drug to market. Although the record
does not show the actual costs Gilead incurred after resuming
TAF development in 2010, it does indicate that
commercialization took an additional five years. During that
five-year period, Gilead conducted multiple phase III clinical
trials involving TAF, the first of which lasted 48 weeks.
The dissent nevertheless suggests that these burdens are
justified because a manufacturer’s patent rights diminish the
ordinary incentives to develop safer alternatives and that
negligence liability is therefore needed as an additional
safeguard. (Dis. opn. of Evans, J., post, at pp. 21–22.) In effect,
the dissent proposes a new negligence rule applicable only to
patent holders, under which patent protection itself justifies
imposing tort liability for delaying the development and
commercialization of alternative drugs. The dissent cites no
authority recognizing such a rule, and we are aware of none.
Congress has already determined that temporary patent
exclusivity is warranted in light of the extraordinary time,
expense, and risk associated with developing new medicines.
Requiring manufacturers, under threat of tort liability, to
accelerate the development and commercialization of products
that may supplant their own patented drugs would
substantially diminish the value of that legislatively created
incentive and risk upsetting the careful balance Congress
struck.
The burdens on manufacturers resulting from the duty are
far greater than those we identified in Verdugo v. Target Corp.
(2014) 59 Cal.4th 312, where we concluded that requiring
Target to purchase, maintain, and train personnel in the use of
automatic external defibrillators at its retail stores “would
49
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
impose considerably more than a minor or minimal burden on a
business establishment.” (Id. at p. 340.) They are also more
substantial than those we described in T.H. v. Novartis
Pharmaceuticals Corp. (2017) 4 Cal.5th 145, where we
recognized that brand-name manufacturers have a duty to warn
not only about the risks of their own drugs, but also about the
risks associated with generic bioequivalents manufactured by
other companies. (Id. at p. 156.) We reasoned that brand-name
manufacturers already have a legal duty to update and
maintain adequate warning labels on brand-name drugs for
known or scientifically knowable side effects associated with
those drugs, and federal law requires generic manufacturers to
copy those labels. (Id. at p. 170.) Hence, “where the brand-name
manufacturer provides an adequate label” for its own drug, “it
necessarily has also fulfilled its duty with respect to the generic
bioequivalent.” (Ibid.)
By contrast, requiring drug manufacturers not to
unreasonably delay bringing alternative drugs to market would
impose substantial burdens. As detailed above, such a duty
would likely compel manufacturers to devote significant time,
money, and resources to developing and commercializing drugs
they might not otherwise prioritize. We therefore conclude that
this factor weighs in favor of recognizing an exception to any
duty of care that might otherwise apply in this context. The
dissent’s reliance on Novartis is misplaced. (Dis. opn. by Evans,
J., post, at p. 25.) The burden of requiring accurate warning
labels for drugs that are already on the market is fundamentally
different from imposing a legal obligation to continue investing
50
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
in and developing alternative drugs based solely on promising
but preliminary clinical trials.
The final policy factor is the availability and cost of
insurance. (Rowland, supra, 69 Cal.2d at p. 113.) In Brown, we
found that public policy considerations weighed against
imposing strict liability for design defects in prescription drugs,
in part because the resulting increase in litigation would likely
raise insurance costs — if insurance were available at all.
(Brown, supra, 44 Cal.3d at pp. 1063–1065.) Although Brown
involved a strict liability claim and, thus, a potentially greater
volume of claims than the negligence theory asserted here,
many of the concerns we identified in Brown apply here as well.
As we explained in Brown, “[t]he possibility that the cost of
insurance and of defending against lawsuits will diminish the
availability and increase the price of pharmaceuticals is far from
theoretical.” (Id. at p. 1064.) Indeed, the added expense of
insuring against liability for failure to commercialize a safer
drug without delay “could place the cost of medication beyond
the reach of those who need it most.” (Id. at p. 1063.)
The Court of Appeal dismissed these concerns on the
ground that the duty would not meaningfully expand drug
manufacturers’ exposure to liability because it applies only
where (1) the manufacturer has already invented or developed
an alternative drug and (2) knows that the alternative “will
avoid significant side effects” associated with an existing drug.
(Gilead Tenofovir Cases, supra, 98 Cal.App.5th at p. 944.) The
court further emphasized that the duty requires only that
manufacturers “act with reasonable care” in their development
and commercialization decisions. (Ibid.) As already discussed,
however, these purported limitations offer little practical
constraint. Drug manufacturers routinely develop backup
51
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
candidates in parallel with lead candidates, and there is
substantial uncertainty as to how to determine whether a
particular backup candidate is sufficiently “invented” or
“developed” to the point that the manufacturer must
commercialize the drug without undue delay. In addition, a
manufacturer cannot know — based on preliminary clinical
data alone — the backup candidate is “safer” than the lead
candidate. Moreover, assessing the “reasonableness” of a
manufacturer’s delay in developing or commercializing an
alternative drug requires a retrospective evaluation of complex
prioritization decisions made with only preliminary clinical
data, such as whether to pursue alternatives that mitigate some
patient risks while exacerbating others, or to allocate resources
toward therapies for unmet medical needs rather than refining
nondefective drugs already on the market. This inquiry poses a
substantial risk of hindsight bias.
For these reasons, the duty would likely result in
significantly greater liability exposure than the Court of Appeal
perceived. The court’s reliance on Carlin, supra, 13 Cal.4th
1104 does not alter this conclusion, because Carlin addressed a
manufacturer’s duty to warn of known or knowable risks
associated with a drug that has already been placed on the
market after completing all clinical trials and obtaining FDA
approval. (See id. at p. 116.) By contrast, the duty asserted here
would expose manufacturers to liability for research,
development, and commercialization decisions made long before
a drug is approved or marketed, and it would require a
backward-looking analysis of scientific judgments and resource-
allocation decisions based on incomplete and evolving data.
52
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
This factor accordingly weighs in favor of recognizing an
exception to the default duty of care.
On balance, the Rowland factors support a judicial
exception to the default duty in this context. Because a drug
manufacturer cannot determine a drug’s safety and efficacy
during the early stages of clinical testing, it cannot reasonably
foresee that any delay in developing and commercializing an
alternative drug will harm users of an existing one. In addition,
the connection between a manufacturer’s decision to delay
developing and commercializing an allegedly safer alternative
drug and a plaintiff’s injury from an existing one is attenuated.
It depends on a chain of uncertain scientific outcomes and
independent decisions by regulators, physicians, and patients.
Rowland’s public policy factors also weigh against
recognizing a duty: Moral blame is not meaningfully implicated,
given the morally neutral and socially valuable reasons that
may underlie drug development decisions and given the lack of
reasonable, practicable ameliorative steps a drug manufacturer
could take. In addition, while a duty might prevent some harms,
it also risks distorting research priorities, discouraging
innovation, and inviting hindsight-based second-guessing of
complex scientific judgments. Finally, the burden such a duty
would impose on drug manufacturers would be substantial. For
these reasons, we conclude that a drug manufacturer has no
duty of care when deciding whether and when to develop and
commercialize an allegedly safer alternative drug.
The dissent observes that we have “never before held that
drug manufacturers are entitled to an exception to the default
negligence duty of care” under Rowland. (Dis. opn. of Evans, J.,
post, at p. 13.) That is unsurprising, given that we have never
53
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
before been presented with plaintiffs’ novel theory that a drug
manufacturer may be liable in negligence for allegedly delaying
the development and commercialization of an alternative drug
to replace a concededly nondefective one.
Ultimately, the dissent’s application of the Rowland
factors turns the analysis into a “heads I win, tails you lose”
exercise, in which every factor necessarily disfavors recognizing
an exception to any duty. As described above, the dissent
asserts that the knowledge component of the foreseeability
inquiry can only be resolved by a fact finder. (Dis. opn. of Evans,
J., post, at p. 15.) But if that were so, this factor would almost
never weigh in favor of an exception to a duty of care. The
dissent similarly concludes that the alleged harm was
foreseeable because Gilead’s decision not to pursue an
alternative drug was “the first step in plaintiffs’ continued
[harm]” and the “root cause of plaintiffs’ alleged harm,” but
ignores all of the intervening decisions and conduct — such as
clinical trial results, FDA approval, or prescribing decisions —
that may determine whether such harm occurs at all. (Id. at
p. 14.) The dissent’s reasoning reduces the foreseeability
inquiry to the allegation of negligence itself: the harm is
foreseeable because the defendant acted negligently. We have
never applied Rowland’s foreseeability factor so broadly.
The dissent’s moral blame analysis suffers from the same
flaw. Its focus on “the reaping of a profit and inherent power
imbalance between plaintiffs and defendants” (dis. opn. of
Evans, J., post, at p. 18) describes virtually every negligence
action against a business. If a company’s status as a for-profit
enterprise necessarily establishes moral blame, that factor too
54
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
will almost always favor liability. The dissent turns application
of the Rowland factors into a vain exercise.
In recognizing an exception to any duty of care a
manufacturer may otherwise owe, we do not minimize the
seriousness of the injuries plaintiffs allegedly suffered. Rather,
our holding reflects the limits of common law negligence as a
tool for regulating the research, development, and
commercialization decisions regarding prescription drugs — an
area already subject to extensive federal oversight. Negligence
law is poorly suited to supervise forward-looking decisions
grounded in uncertain, incomplete, and evolving clinical data
through retrospective assessments of reasonableness.
The dissent repeatedly characterizes today’s decision as
granting pharmaceutical manufacturers “sweeping immunity.”
(Dis. opn. of Evans, J., post, at pp. 1, 29, 32.) It does not. The
FDA rigorously evaluates new drugs before approving them for
commercialization. In addition, plaintiffs injured by FDA-
approved drugs may seek relief under traditional products-
liability principles. Drug manufacturers remain subject to
traditional tort and statutory claims arising from defective
products or wrongful conduct. (Brown, supra, 44 Cal.3d at
p. 1069 & fn. 12) [drug manufacturers are not strictly liable for
design defects but remain liable for negligent design,
manufacturing defects, and failure to warn of known or
reasonably knowable risks].) For example, plaintiffs may try to
establish that a manufacturer violated its duty to market a drug
free from defects by showing that an existing drug was
negligently designed, perhaps due to the availability of an
alternative design whose benefits outweighed the burden of
adopting it. Plaintiffs may also be able to pursue relief under
applicable consumer protection laws, and in appropriate
55
GILEAD TENOFOVIR CASES
Opinion of the Court by Groban, J.
circumstances under common-law fraud principles, by
establishing that a drug manufacturer engaged in unlawful,
unfair, or fraudulent conduct, including by withholding material
information or misleading the public.
What today’s decision declines to do is recognize, for the
first time anywhere, sweeping liability for injuries caused by a
concededly nondefective drug because the manufacturer
allegedly failed to make a different drug available sooner.
Imposing such liability would create substantial burdens and
would risk adverse consequences for pharmaceutical innovation,
public health, and patient safety. For these reasons, we
conclude that drug manufacturers do not owe a duty of care to
users of a nondefective drug when making decisions about
whether and when to commercialize an allegedly safer
alternative drug.
III. DISPOSITION
We reverse the judgment of the Court of Appeal and
remand to that court with directions to issue a writ of mandate
directing the trial court to vacate its order denying summary
judgment and to enter a new order granting summary judgment
for Gilead on all causes of action.
GROBAN, J.
We Concur:
CORRIGAN, J.
LIU, J.
KRUGER, J.
DESAUTELS, J.*
__________________________
* Associate Justice of the Court of Appeal, First Appellate
District, Division Two, assigned by the Chief Justice pursuant to
article VI, section 6 of the California Constitution.
56
GILEAD TENOFOVIR CASES
S283862
Concurring Opinion by Chief Justice Guerrero
It is black letter law that, in order to pursue a negligence
claim against a manufacturer for damages based on the use of
its product, “a plaintiff must prove that a defect caused injury.”
(Merrill v. Navegar, Inc. (2001) 26 Cal.4th 465, 479 (Merrill).)
The majority recognizes this principle, and it expresses
“significant doubts” that plaintiffs’ theory of negligence without
a product defect is viable. (Maj. opn., ante, at p. 13.)
Nonetheless, the majority assumes that a manufacturer owes a
duty of care to its customers and users beyond selling a
nondefective product, and it concludes that an exception to that
duty is warranted under the specific circumstances here. (Id. at
pp. 13–14.)
While I understand the majority’s inclination to avoid the
fundamental questions raised by plaintiffs’ novel theory, I
believe such avoidance is unwarranted. A straightforward
application of this court’s precedents forecloses plaintiffs’ theory
of liability without defect. Fifty years ago, we held that an
essential element of a negligence cause of action against a
manufacturer was that the plaintiff “was injured by a defect in
the product.” (Jiminez v. Sears, Roebuck & Co. (1971) 4 Cal.3d
379, 383 (Jiminez).) We have never deviated from that
requirement. (Merrill, supra, 26 Cal.4th at p. 479.) Plaintiffs’
failure to allege an injury based on a product defect is therefore
fatal to their claim. We need not make any assumptions
regarding a broader duty or consider whether any exception
1
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
might be justified under the factors in Rowland v. Christian
(1968) 69 Cal.2d 108, 113 (Rowland).
Plaintiffs contend that Civil Code section 1714 authorizes
a negligence cause of action without proof of a defective
product.1 Plaintiffs are incorrect. Their contention interprets
section 1714 in a vacuum, and it ignores the entire history of
negligence law in California. Section 1714 was enacted in 1872,
and it has never been interpreted to impose a broad duty of care
on manufacturers. It generally reflects the common law, which
for many decades sharply restricted negligence actions against
manufacturers. The law of products liability freed plaintiffs
from these traditional restrictions, but it did not simply allow
any cause of action sounding in negligence. It required — and
still requires — plaintiffs to prove certain elements in order to
recover. Plaintiffs’ rejection of products liability law does not
allow them to avoid these elements. Instead, it leaves them with
no cause of action at all.
In this case, plaintiffs’ failure to allege a product defect
means they have abandoned any claim that tenofovir disoproxil
fumarate (TDF) is unsafe. In other words, plaintiffs cannot
argue that the likelihood and gravity of any harmful side effects
justifies the burden on defendant Gilead Sciences, Inc. (Gilead)
to commercialize an alternative drug like tenofovir alafenamide
fumarate (TAF). (See Jiminez, supra, 4 Cal.3d at p. 384.)
Gilead acted responsibly by releasing TDF to the market and
saving countless lives. But plaintiffs would have Gilead go
further. They would require Gilead, anytime it releases a safe
and nondefective drug like TDF, to take every reasonable step
1
Subsequent statutory references are to the Civil Code.
2
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
to develop and market additional drugs for the same condition,
or face tort liability for its failure to do so. This unjustified and
dramatic expansion of a manufacturer’s duty to its customers
and users ignores nearly a century of precedent by this court.
While the majority appears sympathetic to this analysis,
it stops short of adopting it as the basis for its holding. Instead,
the majority assumes plaintiffs are correct that manufacturers
owe their customers a broad duty of care, and it holds only that
the specific circumstances here justify an exception to this broad
duty. Although I agree such an exception would be justified, if
manufacturers did have a broad duty of care, I do not support
the majority’s approach. It calls into question fundamental
principles of products liability law, and it creates unnecessary
uncertainty regarding the scope of a manufacturer’s duty in
other industries and in other circumstances. I therefore concur
only in the result reached by the majority, but not its reasoning.
I. SECTION 1714 AND MANUFACTURER LIABILITY
Enacted in 1872, section 1714 provides in relevant part,
“Everyone is responsible, not only for the result of his or her
willful acts, but also for an injury occasioned to another by his
or her want of ordinary care or skill in the management of his
or her property or person, except so far as the latter has,
willfully or by want of ordinary care, brought the injury upon
himself or herself.” (§ 1714, subd. (a).)
Notwithstanding this broad language, manufacturers
historically had no duty in tort to consumers with whom they
had no contractual relationship. Consumers injured by the use
of a product had little recourse against the product’s
manufacturer, unless the consumer dealt with the
manufacturer directly. (See 1 Owen & Davis on Products
3
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
Liability (4th ed. 2026) § 1:10 (Owen & Davis).) The common
law “prohibited negligence actions against ‘remote’
manufacturers, parties with whom injured plaintiffs had no
‘privity of contract.’ ” (Ibid.; see, e.g., Lewis v. Terry (1896)
111 Cal. 39, 44.) As this court has recognized, “the common law,
as a general rule, throws a strong arm of protection around the
manufacturer, warding off claims of third persons, not direct
purchasers, for personal injuries sustained from use of articles
so manufactured and sold by him.” (Kalash v. Los Angeles
Ladder Co. (1934) 1 Cal.2d 229, 231 (Kalash); see Beacon
Residential Community Assn. v. Skidmore, Owings &
Merrill LLP (2014) 59 Cal.4th 568, 574.)
Section 1714 did not itself prompt reconsideration of these
principles. It largely reflected the common law; it did not
subvert it. In one early case, we said, “The common law of
England is declared to be the rule of decision in all the courts of
this state so far as not repugnant to or inconsistent with our
constitution and statutes. [Citation.] The Civil Code was not
designed to embody the whole law of private and civil relations,
rights and duties; it is incomplete and partial; and except in
those instances where its language clearly and unequivocally
discloses an intention to depart from, alter, or abrogate the
common-law rule concerning a particular subject matter, a
section of the code purporting to embody such doctrine or rule
will be construed in the light of common-law decisions on the
same subject.” (In re Estate of Elizalde (1920) 182 Cal. 427, 433;
see Li v. Yellow Cab Co. (1975) 13 Cal.3d 804, 815 (Li).)
Nonetheless, the common law principles governing
manufacturer liability were the subject of further development
in the courts. “By 1900, dissatisfaction with the privity bar had
generated certain exceptions,” including where a plaintiff
4
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
established “ ‘an act of negligence of a manufacturer or vendor
which is imminently dangerous to the life or health of mankind
[in preparing] an article intended to preserve, destroy, or affect
human life.’ ” (Owen & Davis, supra, § 1:10.)
In Kalash, this court adopted the broad exception
articulated in the seminal case of MacPherson v. Buick Motor
Co. (N.Y. 1916) 111 N.E. 1050, 1053 (MacPherson): “ ‘[The
exception] is not limited to poisons, explosives, and things of like
nature, to things which in their normal operation are
implements of destruction. If the nature of a thing is such that
it is reasonably certain to place life and limb in peril when
negligently made, it is then a thing of danger. Its nature gives
warning of the consequences to be expected. If to the element of
danger there is added knowledge that the thing will be used by
persons other than the purchaser, and used without new tests,
then irrespective of contract, the manufacturer of this thing of
danger is under a duty to make it carefully.’ ” (Kalash, supra,
1 Cal.2d at pp. 231–232.)
This duty did not arise from a contractual relationship. It
was implied by law. As Kalash recognized, “ ‘We have put aside
the notion that the duty to safeguard life and limb, when the
consequences of negligence may be foreseen, grows out of
contract and nothing else. We have put the source of the
obligation where it ought to be. We have put its source in the
law.’ ” (Kalash, supra, 1 Cal.2d at p. 232, quoting MacPherson,
supra, 111 N.E. at p. 1053.) Although the precise source of the
duty was variously identified, we have held the “true reason for
the rule” is found in the Civil Code, including section 1714.
(Dahms v. General Elevator Co. (1932) 214 Cal. 733, 739
(Dahms).)
5
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
Thus, from its origins, the law of products liability arose
not in opposition to section 1714, but as an interpretation of its
broad foundational principle, as “developed . . . from the
common law.” (Kentucky Fried Chicken of Cal., Inc. v. Superior
Court (1997) 14 Cal.4th 814, 828 (Kentucky Fried Chicken).)
Further development resulted in the modern law of
products liability. The exception for a product “ ‘negligently
made’ ” (Kalash, supra, 1 Cal.2d at p. 231) was split into two
analytically distinct concepts: negligent manufacturing and
negligent design. (Merrill, supra, 26 Cal.4th at p. 479; see
Rest.2d Torts, §§ 395 [negligent manufacturing], 398 [negligent
design].) Additionally, courts recognized that a manufacturer
may be liable for injuries caused by its negligent failure to warn
of known or reasonably knowable dangers in its products.
(Stevens v. Parke, Davis & Co. (1973) 9 Cal.3d 51, 64 (Stevens);
see Webb v. Special Electric Co., Inc. (2016) 63 Cal.4th 167, 181
(Webb); Rest.2d Torts, § 388.)
Thus, almost 40 years after Kalash, this court observed,
“[O]ver the years a considerable body of law has been developed
as to negligence permitting definitive instructions based upon
tested and settled principles.” (Jiminez, supra, 4 Cal.3d at
p. 384.) For example, we stated in Pike v. Frank G. Hough Co.
(1970) 2 Cal.3d 465 that the relevant principle of negligent
design was settled: “The duty of a manufacturer with respect to
the design of products placed on the market is defined in the
Restatement Second of Torts, section 398: ‘A manufacturer of a
chattel made under a plan or design which makes it dangerous
for the uses for which it is manufactured is subject to liability to
others whom he should expect to use the chattel or to be
endangered by its probable use for physical harm caused by his
failure to exercise reasonable care in the adoption of a safe plan
6
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
or design.’ Thus, the manufacturer must use reasonable care ‘to
so design his product as to make it not accident-proof, but safe
for the use for which it was [sic] intended.’ [Citation.] What is
‘reasonable care,’ of course, varies with the facts of each case,
but it involves a balancing of the likelihood of harm to be
expected from a machine with a given design and the gravity of
harm if it happens against the burden of the precaution which
would be effective to avoid the harm.” (Id. at p. 470; accord,
Jiminez, at p. 384.)
To shorthand these principles, courts have often described
a negligently designed or negligently manufactured product, or
a product accompanied by inadequate warnings, as “defective”
or suffering from a specific “defect.” (Merrill, supra, 26 Cal.4th
at pp. 479–480; see, e.g., Webb, supra, 63 Cal.4th at p. 181;
Jiminez, supra, 4 Cal.3d at p. 383; Kalash, supra, 1 Cal.2d at
p. 231; Dahms, supra, 214 Cal. at pp. 738, 742; MacPherson,
supra, 111 N.E. at pp. 1051, 1053.) The law of products liability,
then, has been described as “ ‘the area of the law involving the
liability of those who supply goods or products for the use of
others to purchasers, users, and bystanders for losses of various
kinds resulting from so-called defects in those products.’ ”
(Merrill, at p. 478.)
The defect requirement is crucial. “[D]efect is the
conceptual linchpin that holds products liability law together; a
system of liability without defect is beyond the capacity of courts
to implement.” (Henderson & Twerski, Closing the American
Products Liability Frontier: The Rejection of Liability Without
Defect (1991) 66 N.Y.U. L.Rev. 1263, 1267.)
7
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
II. SECTION 1714 AS A SEPARATE DUTY
Plaintiffs argue they can pursue a negligence action for
injuries caused by TDF without alleging a product defect, based
on the broad language of section 1714. They point out that we
have described section 1714 as articulating the “general rule in
California . . . that ‘[e]veryone is responsible . . . for an injury
occasioned to another by his or her want of ordinary care or skill
in the management of his or her property or person . . . .”
(Cabral v. Ralphs Grocery Co. (2011) 51 Cal.4th 764, 771
(Cabral).)
The flaw in plaintiffs’ argument is that the general duty
in section 1714 has never been held sufficient to impose liability
on a manufacturer for injuries caused by the use of a
manufacturer’s product. Instead, courts developed the law of
products liability to hold manufacturers liable, but only where
plaintiffs can allege a defect caused their injuries. If section
1714 itself authorized a plaintiff to sue in negligence without
proof of a defect, it would have been unnecessary for courts to
develop the principles of products liability described above, and
it would be unnecessary for plaintiffs to adhere to them. In
other words, plaintiffs ask this court to simply recognize that
section 1714 has always authorized a negligence claim against
a manufacturer for damages caused by the use of its product,
even in the absence of a defect, and “to brush aside all of the
misguided decisions which have concluded otherwise up to the
present day.” (Li, supra, 13 Cal.3d at p. 817.) This “arresting
contention” refutes itself. (Ibid.)
As noted, section 1714 “states a civil law principle that is
the foundation of our negligence law.” (Kentucky Fried Chicken,
supra, 14 Cal.4th at p. 828.) But it is merely the foundation. It
8
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
does not supplant the more specific standards that “may be
established by judicial decision, statute or ordinance.” (Id. at
p. 824.) “ ‘[T]he proper conduct of a reasonable person under
particular situations may become settled by judicial decision or
be prescribed by statute or ordinance,’ ” and in that situation the
general duty of care no longer applies. (Ramirez v. Plough, Inc.
(1993) 6 Cal.4th 539, 547; see Rest.2d Torts, § 285.) Here, our
prior precedents have defined the scope of a manufacturer’s duty
of care to avoid harm caused by the use of its products. Plaintiffs
cannot rely on section 1714 to circumvent those precedents.2
We confronted a similar situation in Parsons v. Crown
Disposal Co. (1997) 15 Cal.4th 456, 460 (Parsons), where a
plaintiff brought a general negligence claim against a defendant
based on injuries sustained when his horse was startled by
defendant’s trash collection activities. We observed that “the
2
In numerous other contexts, we have recognized that
section 1714’s broad statement of duty cannot simply be applied
to every plaintiff and defendant. “A case in point is liability in
negligence for purely economic losses, which is ‘the exception,
not the rule’ under our precedents. [Citation.] And that holds
true even though . . . section 1714 does not, by its terms,
‘distinguish among injuries to one’s person, one’s property or
one’s financial interests.’ ” (Southern California Gas Leak Cases
(2019) 7 Cal.5th 391, 400; see Bily v. Arthur Young & Co. (1992)
3 Cal.4th 370, 397.) Other examples include the duty owed to
coparticipants in active sports, which is limited to intentional
injuries to another participant or “conduct that is so reckless as
to be totally outside the range of the ordinary activity involved
in the sport” (Knight v. Jewett (1992) 3 Cal.4th 296, 320) and the
duty to warn third parties of danger, which is limited to
circumstances where the plaintiff can establish a special
relationship between the defendant and the plaintiff or the
defendant and the third party (Tarasoff v. Regents of University
of California (1976) 17 Cal.3d 425, 435–436).
9
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
nature and scope of defendant’s duty in these circumstances is
established by the considerable line of authority addressing the
question of a defendant’s potential liability for injuries resulting
from the frightening of a horse.” (Id. at p. 462.) At common law,
“courts developed a remarkably uniform rule, holding that a
plaintiff whose horse ‘shied’ or ‘spooked’ and caused damage
because of the noise, sight, or odor caused by the defendant’s
regular and necessary conduct, cannot state a cause of action for
negligence, because the defendant in such a case has breached
no duty of care.” (Id. at p. 466.) “In each category of case,
however, the courts recognized ‘exceptions’ to the general rule of
nonliability.” (Id. at p. 469.) These exceptions allowed liability
where the plaintiff could establish, for example, that a
defendant operated a machine “in a careless or imprudent
manner, or causes noises or emissions unnecessary to the
regular operation of the machine.” (Id. at pp. 469–470.)3
In Parsons, we recognized the general duty of reasonable
care embodied by section 1714. (Parsons, supra, 15 Cal.4th at
p. 472.) But we did not apply section 1714 in a vacuum. We
examined “the existence and scope of duty” “[w]ith these
principles and this history in mind.” (Parsons, at pp. 473, 472.)
In the end, we saw “no basis for treating defendant’s garbage
collection truck differently from the various machines and
3
Courts recognized other exceptions where “the defendant
fails to take reasonable protective actions after it knows that the
plaintiff’s horse actually has become frightened,” “the defendant
or its employees conduct its machinery in an unnecessary or
malicious fashion designed to cause fright,” or “the defendant
violates a safety statute designed to protect the class of which
the plaintiff is a member.” (Parsons, supra, 15 Cal.4th at
p. 470.)
10
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
devices discussed [in prior cases], or for increasing the burden
on machine operators over what was considered reasonable in
an earlier age.” (Id. at p. 474.) We “reject[ed] plaintiff’s
assertion that defendant owed an expansive duty to guard
against frightening horses” but “affirm[ed] that defendant was
obligated to conduct itself in accordance with the limited
common law duty” established previously. (Id. at p. 477.)
A similar approach governs this case. Plaintiffs cannot
simply invoke section 1714 to avoid the common law principles
governing their cause of action against Gilead. (See Parsons,
supra, 15 Cal.4th at p. 461 [criticizing the lower court for
adopting “a broad, expansive duty . . . that takes no account of
the established authority” defining the defendant’s duty of
care].) Instead, plaintiffs must justify a departure from common
law principles based on “ ‘ “ ‘the sum total of those
considerations of policy which lead the law to say that the
particular plaintiff is entitled to protection.’ ” ’ ” (Id. at p. 472.)
“Duty is not universal; not every defendant owes every plaintiff
a duty of care. A duty exists only if ‘ “the plaintiff’s interests are
entitled to legal protection against the defendant’s conduct.” ’ ”
(Brown v. USA Taekwondo (2021) 11 Cal.5th 204, 213.)
Plaintiffs do not attempt to justify a departure from
traditional principles of products liability. Instead, they claim
that section 1714 itself accomplishes this result. They are
mistaken, for reasons I have already explained.
Plaintiffs argue that Rowland, supra, 69 Cal.2d 108,
supports their view of section 1714. In Rowland, this court
abandoned the common law approach to a landowner’s duties,
which was based on the status of the injured plaintiff as a
licensee, an invitee, or a trespasser. (Rowland, at pp. 113–114.)
11
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
By the time Rowland was decided, the common law approach
had become riddled with complex and confusing exceptions. (Id.
at pp. 115–116.) We remarked, “Whatever may have been the
historical justifications for the common law distinctions, it is
clear that those distinctions are not justified in the light of our
modern society and that the complexity and confusion which has
arisen is not due to difficulty in applying the original common
law rules — they are all too easy to apply in their original
formulation — but is due to the attempts to apply just rules in
our modern society within the ancient terminology.” (Id. at
p. 117.) We concluded that adhering to these common law
distinctions “can only lead to injustice or, if we are to avoid
injustice, further fictions with the resulting complexity and
confusion.” (Id. at p. 119.)
Thus, even in Rowland, we did not treat section 1714 as
solely determinative of a defendant’s duty. We examined the
traditional common law approach to determine whether it
continued to be persuasive in light of the “basic policy” of the
state embodied in section 1714. (Rowland, supra, 69 Cal.2d at
p. 118.) “The factors which may in particular cases warrant
departure from this fundamental principle do not warrant the
wholesale immunities resulting from the common law
classifications . . . .” (Id. at p. 119.) We did not read section
1714 in a vacuum, and we did not hold that section 1714 itself
allowed any plaintiff to sue any defendant for negligence,
regardless of any common law limitation on a defendant’s duty
under specific circumstances.4
4
In addition, at issue in Rowland were “ancient concepts as
to the liability of the occupier of land” (Rowland, supra,
12
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
Plaintiffs further argue that, even in the realm of products
liability law, proof of a product defect is not always required to
support a negligence action against a manufacturer. This
argument is refuted by the history of products liability law
discussed above. Moreover, the cases cited by plaintiffs do not
support dispensing with the defect requirement.
For example, in Stevens, supra, 9 Cal.3d 51, we concluded
that plaintiffs had presented sufficient evidence to sustain a
claim against a drug manufacturer for negligent failure to warn.
(Id. at p. 64.) The defect in that case was the incorrect or
inadequate information provided to medical professionals about
the product. (Id. at pp. 64–65.) Evidence of the manufacturer’s
overpromotion, along with its inadequate product warnings,
established a breach of the manufacturer’s specific duty in
products liability “ ‘to exercise reasonable care to inform [users
or their physicians] of [a product’s] dangerous condition or of the
facts which make it likely to be dangerous.’ ” (Id. at p. 64; see
id. at pp. 66–67.) We did not suggest that proof of a product
defect was unnecessary. To the contrary, we have confirmed
that liability for a manufacturer’s negligent failure to warn
69 Cal.2d at p. 119) that were no longer “justified in the light of
our modern society” (id. at p. 117). The law of products liability,
by contrast, arose in response to the commercial realities of
modern society. (See Owen & Davis, supra, § 1:10 [discussing
legal developments after “manufacturing and retailing
functions separated in the 1800s during the course of
industrialization”].) Similarly, the “ancient concepts”
abandoned in Rowland stood in opposition to the general
principle in section 1714. (Rowland, at p. 119.) The law of
products liability, by contrast, explicitly furthers its purpose.
(Dahms, supra, 214 Cal. at p. 740.)
13
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
depends on proof of a “ ‘ “warning defect.” ’ ” (Webb, supra,
63 Cal.4th at p. 181.)
Similarly, in Hasson v. Ford Motor Co. (1977) 19 Cal.3d
530, 543–544, we upheld a jury verdict against a manufacturer
based on negligent failure to warn. In response to a special
interrogatory, the jury had found “that there had been no ‘defect’
in the [plaintiffs’] vehicle ‘at the time it was manufactured and
sold.’ ” (Id. at p. 539.) The manufacturer argued that this
finding was fatally inconsistent with the jury’s finding of
negligence. (Id. at p. 540.) We disagreed, but not because proof
of a defective product was unnecessary. Instead, we held, “The
jury could reasonably believe that the
interrogatory . . . intended to exclude consideration of any
‘defect’ based on a failure to warn.” (Id. at p. 543.)
Plaintiffs rely heavily on Mexicali Rose v. Superior Court
(1992) 1 Cal.4th 617, but it has no application here. There, we
approved of a specific duty of care, i.e., a restaurant must
“exercise reasonable care in the preparation of [its] food.” (Id.
at p. 633.) We recognized that a restaurant patron injured by
the presence of a natural substance in food (there, a chicken
bone) may maintain an action based on negligent preparation
even though the food is not considered defective. (Ibid.) But our
recognition stemmed from the unique context of food, which may
have qualities or components that are both (1) natural and
inherent in the food itself and (2) potentially harmful when
included in a restaurant dish. It would not make sense to call
such a natural quality or component a “defect” in the food.
Contrary to the view of the Court of Appeal below, Mexicali Rose
does not stand for the proposition that “a plaintiff may recover
under the doctrine of negligence for harm caused by a product
otherwise subject to the doctrine of strict liability,
14
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
notwithstanding the plaintiff’s inability to prove a product
defect.” (Gilead Tenofovir Cases (2024) 98 Cal.App.5th 911,
926.) A food product that contains only natural substances is
not subject to the doctrine of strict liability because “the food
cannot be determined unfit or defective.” (Mexicali Rose, at
p. 633.) In any event, our endorsement of a specific cause of
action against restaurants based on their negligence resulting
in nondefective (but still harmful) food does not say anything
about the defect requirement in the context of nonfood products.
Finally, plaintiffs contend that the defect requirement
improperly imports an element from the doctrine of strict
liability into a negligence cause of action. Again, plaintiffs are
incorrect. As discussed, “under either a negligence or a strict
liability theory of products liability, to recover from a
manufacturer, a plaintiff must prove that a defect caused
injury.” (Merrill, supra, 26 Cal.4th at p. 479.) In negligence
actions, we have consistently identified a product defect as a
required element, even before the emergence of the doctrine of
strict products liability. (See, e.g., Kalash, supra, 1 Cal.2d at
p. 231; Dahms, supra, 214 Cal. at pp. 738, 742; see also
MacPherson, supra, 111 N.E. at pp. 1051, 1053.) Plaintiffs
further contend that these early cases refer to product defects
only because plaintiffs alleged a defect, not because a product
defect was required. They maintain that, at common law,
remote manufacturers could be held liable in general negligence
for injuries caused by a product, without proof of a product
defect. Plaintiffs cite no authority for this contention, and it flies
in the face of a century of precedent establishing that remote
manufacturers were not liable except where plaintiffs could
prove a product was defective.
15
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
In sum, neither section 1714 nor our precedents provide
any basis for plaintiffs’ theory of negligence without proof of
defect. In order to maintain a negligence action against a
manufacturer based on injuries caused by the manufacturer’s
product, a plaintiff must prove the injuries were caused by a
defective product. A plaintiff may not avoid this essential
element by invoking section 1714.
III. THE MERITS OF A SEPARATE DUTY
Because they contend that section 1714 already imposes a
broad duty of reasonable care on manufacturers, plaintiffs do
not attempt to justify its creation. Nonetheless, it is easy to see
why expanding a manufacturer’s tort liability to encompass
every step leading to a manufacturer’s decision not to develop
an alternative product would be unwarranted.
Existing law already requires manufacturers to consider
reasonable alternatives to their products. For example, to avoid
liability for a negligent design defect, a manufacturer must act
reasonably in balancing the likelihood and gravity of the harm
to be expected from the existing design against the burden of
adopting an alternative design that would avoid the harm.
(Merrill, supra, 26 Cal.4th at p. 479.) Indeed, the manufacturer
would also be strictly liable “ ‘either (1) if the product has failed
to perform as safely as an ordinary consumer would expect when
used in an intended or reasonably foreseeable manner, or
(2) if . . . the benefits of the challenged design do not outweigh
the risk of danger inherent in such design.’ ” (Ibid.) As to the
latter, if a plaintiff can show that the design of a product caused
injury, the plaintiff is entitled to relief unless the manufacturer
can “ ‘establish, in light of the relevant factors, that, on balance,
the benefits of the challenged design outweigh the risk of danger
16
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
inherent in such design.’ ” (Kim v. Toyota Motor Corp. (2018)
6 Cal.5th 21, 30.) Relevant considerations include “ ‘the gravity
of the danger posed by the challenged design, the likelihood that
such danger would occur, the mechanical feasibility of a safer
alternative design, the financial cost of an improved design, and
the adverse consequences to the product and to the consumer
that would result from an alternative design.’ ” (Ibid.) Thus,
under existing law, a manufacturer may be liable “ ‘ “when the
foreseeable risks of harm posed by the product could have been
reduced or avoided by the adoption of a reasonable alternative
design by the seller or other distributor.” ’ ” (Trejo v. Johnson &
Johnson (2017) 13 Cal.App.5th 110, 142; see Rest.3d Torts,
Products Liability, § 2(b).)
These principles ensure that manufacturers’ products are
reasonably safe and nondefective. But plaintiffs’ broad duty of
care would require manufacturers to go further. Every
manufacturer could face liability for its failure to develop
alternative products, or do so promptly, even where existing
principles of products liability would not require the adoption of
an alternative design at all, because the likelihood and gravity
of any potential injuries would not justify the burden on a
manufacturer to adopt an alternative design that would avoid
the harm. (See Jiminez, supra, 4 Cal.3d at p. 384.) A
manufacturer’s liability would no longer turn on the safety of
the products being sold, but instead on each business decision
made in the course of alternative product development. Such a
dramatic expansion of tort liability would be unprecedented and
unjustified.
As a doctrinal matter, we examine certain considerations
“to determine the existence and scope of duty,” including “ ‘the
foreseeability of harm to the plaintiff, the degree of certainty
17
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
that the plaintiff suffered injury, the closeness of the connection
between the defendant’s conduct and the injury suffered, the
moral blame attached to the defendant’s conduct, the policy of
preventing future harm, the extent of the burden to the
defendant and consequences to the community of imposing a
duty to exercise care with resulting liability for breach, and the
availability, cost, and prevalence of insurance for the risk
involved.’ ” (Parsons, supra, 15 Cal.4th at p. 473.) Although
these factors mirror the Rowland factors considered by the
majority, they serve a different purpose in this context. The
majority considers these factors to determine whether to
recognize an exception to its assumed duty of care. (Maj. opn.,
ante, at p. 22.) But because our precedents already define a
manufacturer’s duty of care in this circumstance, an exception
is not at issue. Instead, we must consider these factors to
determine whether our precedents should be reconsidered and a
manufacturer’s duty expanded to encompass the broad duty
plaintiffs seek to invoke. (See Parsons, at pp. 472–475.)
Because plaintiffs contend section 1714 imposes a broad duty on
all manufacturers, and it is not limited to the pharmaceutical
industry, I consider these factors at a high level of generality.5
5
In their briefing, plaintiffs emphasize their specific
allegations against Gilead, including that Gilead knew TAF was
safer than TDF and that Gilead intentionally delayed its
development of TAF in order to maximize its profits. These
evidentiary allegations are relevant to show breach of a duty,
but they do not aid a court in determining whether a duty exists
in the first place. We found analogous details relating to the
circumstances of an automobile accident irrelevant in Cabral,
supra, 51 Cal.4th at page 774. We explained, “On the duty
question that is presented here, the factual details of the
accident are not of central importance. That [the driver] parked
18
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
To begin, the harm to plaintiffs in this context is not
foreseeable. “ ‘[A] court’s task — in determining “duty” — is not
to decide whether a particular plaintiff’s injury was reasonably
foreseeable in light of a particular defendant’s conduct, but
rather to evaluate more generally whether the category of
negligent conduct at issue is sufficiently likely to result in the
kind of harm experienced that liability may appropriately be
imposed . . . .’ ” (Parsons, supra, 15 Cal.4th at p. 476.) It is not
a factual assessment of foreseeability; it is a “very different and
normative inquiry.” (Ibid.)
The conduct at issue here is a manufacturer’s decision not
to release an alternative product. Innumerable steps might lead
to such a decision, just as innumerable alternatives might exist
to a product actually sold. As a categorical matter, any given
decision not to release an alternative product, where the existing
product is reasonably safe and nondefective, does not itself carry
a sufficient risk of foreseeable harm. The negligent failure to
release an alternative product is not inherently harmful. It
constitutes no affirmative action. Indeed, it is only those rare
alternative products that would avoid harm, yet not rise to the
level of a reasonable alternative design under products liability
16 feet from the outermost traffic lane, rather than six feet or
26 feet; that parking for emergencies was permitted in the dirt
area he chose; that [the plaintiff] likely left the highway because
he fell asleep or because of some unknown adverse health event,
rather than from distraction or even intoxication — none of
these are critical to whether [the driver] owed [the plaintiff] a
duty of ordinary care. These facts may have been important to
the jury’s determinations of negligence, causation and
comparative fault, but on duty California law looks to the entire
‘category of negligent conduct,’ not to particular parties in a
narrowly defined set of circumstances.” (Ibid., italics omitted.)
19
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
law, that would be encompassed within plaintiffs’ broad duty of
care. In many industries, a manufacturer will have difficulty
understanding the comparative impacts on safety of its existing
products and yet-to-be released alternative products, including
whether an alternative product would be safer along certain
dimensions but less safe along others. This difficulty increases
greatly if a manufacturer must foresee how each specific
development decision will affect the ultimate product’s safety.
Even in industries where the level of scientific inquiry and
knowledge is fairly high, it would still be difficult for
manufacturers to gain sufficient knowledge to accurately
foresee the risk of harm. (See maj. opn., ante, at pp. 24–26.)
Similarly, while in this context we can be relatively
certain that a plaintiff suffered harm, the connection between
defendant’s conduct and the harm is quite attenuated. “[W]here
the injury suffered is connected only distantly and indirectly to
the defendant’s negligent act, the risk of that type of injury from
the category of negligent conduct at issue is likely to be deemed
unforeseeable.” (Cabral, supra, 51 Cal.4th at p. 779.) For
example, where a nondefective product causes injury to a
plaintiff, it is far from clear whether the harm would have been
avoided by the release of an alternative product. The
availability of consumer products depends on complicated
distribution networks, and retail outlets may not even stock an
alternative product. Even if available, the decision by a
consumer to use a product may rest on any number of factors,
and it may be pure chance whether a user chooses to use the
original product, an alternative product, or some other
manufacturer’s product. Given this unpredictable and
contingent path, the harm that a user might suffer from the use
of a nondefective product is quite remote from a manufacturer’s
20
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
decision to commercialize an alternative product. (See
Vasilenko v. Grace Family Church (2017) 3 Cal.5th 1077, 1086
[harm is remote where it results from “the confluence” of
decisions by various actors].)
Moreover, even if the risk of harm were sufficiently
foreseeable, “ ‘we will not treat the mere presence’ of such a
finding, ‘standing alone,’ as imposing on defendant a duty to
guard against injuries to plaintiff. [Citation.] As we have
observed, ‘social policy must at some point intervene to delimit
liability’ even for foreseeable injury [citation], and ‘policy
considerations may dictate a cause of action should not be
sanctioned no matter how foreseeable the risk.’ ” (Parsons,
supra, 15 Cal.4th at p. 476.) These policy considerations include
the moral blame attached to the category of negligent conduct,
the societal consequences of imposing a duty of care in this
context, and the cost and availability of insurance for the risk
involved.
The moral blame that attaches to a manufacturer’s
negligent decision to forgo release of an alternative product,
when the product actually released is reasonably safe and
nondefective, is fairly low. While ideally every product would be
as safe as possible, and even safer alternatives would be freely
available, a mistake that makes a manufacturer’s portfolio of
products somewhat less safe — but still safe and nondefective —
is not highly morally blameworthy, when compared with other
categories of negligent conduct.
While “the ‘policy of preventing future harm is ordinarily
served, in tort law, by imposing the costs of negligent conduct
upon those responsible,’ ” we must consider in this context “both
the positive and the negative societal consequences of
21
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
recognizing a tort duty.” (Kuciemba v. Victory Woodworks, Inc.
(2023) 14 Cal.5th 993, 1026 (Kuciemba).) Here, the negative
consequences of plaintiffs’ broad duty would be significant.
Given the difficulty of understanding the comparative risks and
benefits of the original product and any unreleased alternative
product, a prudent manufacturer would be incentivized to
release many more products than necessary, dramatically
increasing the costs ultimately borne by consumers.
Alternatively, a prudent manufacturer would be incentivized to
release fewer new products that benefit new classes of
consumers or users, since each released product — even if
reasonably safe and nondefective — risks exposing the
manufacturer to liability for its failure to release alternative
products. Overall, manufacturers would spend significantly
more time and effort releasing alternative products rather than
developing new innovations. (See maj. opn., ante, at pp. 37–50.)
Further, plaintiffs’ broad duty would increase compliance
and litigation costs, since every development decision could lead
to tort liability. Since manufacturers might be exposed to
liability for products they did not release, every possible
alternative product (i.e., every development path not taken)
would have to be scrutinized in the same manner as a product
actually released. Moreover, by imposing liability on
manufacturers who released only safe and nondefective
products, plaintiffs’ broad duty would threaten to impose
liability “ ‘out of proportion to fault,’ ” contrary to the goal of
California’s tort law. (Kuciemba, supra, 14 Cal.5th at p. 1030.)
It is fundamental that a manufacturer “should not be
responsible for all injuries involving the use of its products.”
(Cronin v. J.B.E. Olson Corp. (1972) 8 Cal.3d 121, 133.) A
manufacturer should be liable only “for all injuries proximately
22
GILEAD TENOFOVIR CASES
Guerrero, C. J., concurring
caused by any of its products which are adjudged ‘defective.’ ”
(Id. at pp. 133–134.)
Similarly, insuring against the liability imposed by
plaintiffs’ alleged duty would be uncertain and costly. It will be
difficult for insurers to predict a manufacturer’s exposure to
liability based on products the manufacturer has not developed
or released. Insurers will therefore increase their premiums to
cover the risk involved, if insurance for such unknown risks is
even available.
In sum, plaintiffs have not shown that manufacturers’
existing duties under the law of products liability should be
expanded to encompass a broad duty of reasonable care with
respect to all development decisions. (See Parsons, supra,
15 Cal.4th at p. 478.) A manufacturer that sells a reasonably
safe and nondefective product does not owe its existing
customers and users a duty to develop a reasonable alternative
product.
IV. CONCLUSION
Because plaintiffs’ general negligence cause of action rests
on a flawed premise, and they have not attempted to satisfy the
essential elements of a negligence claim under established
products liability law, Gilead was entitled to summary
judgment. Because the majority reaches the same result, but by
different reasoning, I concur in the result only.
GUERRERO, C. J.
23
GILEAD TENOFOVIR CASES
S283862
Concurring Opinion by Justice Kruger
As the majority opinion explains, established precedent
limits a manufacturer’s liability in negligence for consumer
injuries to cases in which plaintiffs prove a product defect. This
precedent forecloses plaintiffs’ theory of negligence, under which
a manufacturer can be liable for selling a sufficiently safe, and
thus defect-free, product if a jury later determines that the
manufacturer should have spent its time and resources
developing an even safer alternative product.
While I share the Chief Justice’s skepticism that Civil
Code section 1714 overrides this defect requirement, I am
nonetheless willing to assume with the majority that there
might be some conceivable theory of negligence, not yet
identified in our cases, based on the sale of a nondefective
product. Even assuming that manufacturers owe consumers a
general duty of care apart from the marketing of defect-free
products, I agree with the majority that foreseeability and
public policy considerations would warrant an exception in this
context. I write separately to underscore that the majority
makes the threshold assumption for the sake of argument only.
Together, the majority opinion and Chief Justice Guerrero’s
concurrence illustrate not only the tension between plaintiffs’
proposed duty of care and established case law but also the
significant administrability and public policy concerns that, as
the majority says, “may ultimately undermine, rather than
advance, public health and safety.” (Maj. opn., ante, at p. 4.)
1
GILEAD TENOFOVIR CASES
Kruger, J., concurring
With these observations, I concur in the majority opinion.
KRUGER, J.
We Concur:
CORRIGAN, J.
DESAUTELS, J. *
*
Associate Justice of the Court of Appeal, First Appellate
District, Division Two, assigned by the Chief Justice pursuant
to article VI, section 6 of the California Constitution.
2
GILEAD TENOFOVIR CASES
S283862
Dissenting Opinion by Justice Evans
Today, the majority provides pharmaceutical
manufacturers with sweeping immunity from negligence
liability, no matter how unreasonably they may act or how much
serious and avoidable harm they may cause consumers by
intentionally delaying the commercialization of safer drugs.
Pharmaceutical manufacturers benefit society by producing life-
saving drugs. In exchange, they receive special
accommodations, including exclusive patent protections and
exemptions from strict products liability. Consequently, they
reap substantial profits while serving an important public
function. Given these existing protections, I do not believe drug
manufacturers are further entitled to an exception to Civil Code
section 1714’s1 default requirement that everyone must exercise
ordinary care. Like other entities, drug manufacturers should
be held liable for injuries caused when they act negligently or
engage in willful misconduct.
The default duty of care should attach when drug
manufacturers make decisions about whether and when to
commercialize a drug known to be significantly safer than and
as effective as an existing one. This conclusion aligns with
traditional tort law principles, our precedent, and the realities
of the pharmaceutical industry. I therefore respectfully dissent.
1
All subsequent references are to the Civil Code unless
otherwise specified.
1
GILEAD TENOFOVIR CASES
Evans, J., dissenting
I. FACTUAL BACKGROUND
Where, as here, “a motion for summary judgment is used
to test whether the complaint states a cause of action, the court
will apply the rule applicable to demurrers and accept the
allegations of the complaint as true.” (American Airlines, Inc. v.
County of San Mateo (1996) 12 Cal.4th 1110, 1118.) The
plaintiffs’ complaint thus offers the relevant facts. In 1991,
Gilead Sciences, Inc. (Gilead), obtained an exclusive license to
develop, manufacture, and sell tenofovir, an antiretroviral
medication, as a treatment for HIV/AIDs. Tenofovir produces
rapid and severe decline in kidney function when injected
directly into the body, but it is ineffective when consumed orally.
Gilead therefore developed an alternative form of the chemical,
tenofovir disoproxil fumarate (TDF) that could be a usable
medication when administered orally. In 1997, Gilead applied
for approval to begin human clinical testing of TDF, which was
approved by the Food and Drug Administration (FDA) in
October 2001.
Despite the alternative form, TDF still has the potential
for harmful side effects. At some point in TDF’s development,
Gilead developed a different form of tenofovir, tenofovir
alafenamide fumarate (TAF), which Gilead formulated to be as
effective as TDF at a lower dosage than TDF.
In November 2001, Gilead obtained permission to begin
clinical testing of TAF. It conducted a combined phase I/II trial
to compare the safety and effectiveness of TAF relative to TDF.
Plaintiffs allege that Gilead’s trial confirmed what preclinical
research had shown — that TAF was just as effective as TDF
but with lowered risks of renal, bone, and tooth injuries. As a
result of this research, plaintiffs allege that Gilead knew that
2
GILEAD TENOFOVIR CASES
Evans, J., dissenting
TAF was as efficacious as and less toxic to kidneys and bones
than TDF.
Despite this knowledge, Gilead discontinued TAF
development in 2004, allegedly because it believed TAF would
cannibalize TDF’s sales. According to plaintiffs’ complaint,
Gilead chose to wait until TDF’s patent expired before bringing
TAF to market in order to maximize profits on both drugs.
Gilead resumed testing of TAF in 2010 and obtained FDA
approval in 2015. Gilead’s TDF patent expired in 2017.
Plaintiffs presented evidence that, from 2001 through
2021, Gilead generated $101.8 billion in revenue from the sale
of its tenofovir-based HIV drugs. TDF’s annual revenues
peaked at $10.7 billion dollars in 2016, declining every year
after. As TDF revenue declined, TAF revenues increased every
year from its release in 2015. In total, Gilead generated $53.7
billion in revenue from the sale of TAF-based drugs from its
release in 2015 through 2021.
Plaintiffs do not contend that TDF is defective or that
Gilead should have withdrawn it from the market. Instead,
plaintiffs assert that Gilead unreasonably delayed bringing TAF
to market — which they believe could have been commercialized
as early as 2006 — and that this delay deprived them, for many
years, of the option to take a safer drug. Plaintiffs allege that,
as a result, they suffered serious renal, bone, or tooth injuries
from taking TDF that could have been avoided had Gilead made
TAF available sooner.
Gilead moved for summary judgment on plaintiffs’ claims
for negligence and fraudulent concealment. The trial court
denied Gilead’s motion on both counts. The Court of Appeal
reversed the denial of summary adjudication as to the
fraudulent concealment claim but affirmed the denial of
3
GILEAD TENOFOVIR CASES
Evans, J., dissenting
summary adjudication of plaintiffs’ negligence claim. The court
reasoned that “the legal duty of a manufacturer to exercise
reasonable care can, in appropriate circumstances, extend
beyond the duty not to market a defective product.” (Gilead
Tenofovir Cases (2024) 98 Cal.App.5th 911, 917 (Gilead).) The
court further concluded that, on the record before it, Gilead
could not justify a narrower exception to the general duty of
care. (Ibid.)
II. LEGAL BACKGROUND
Section 1714 “establishes a default duty requiring all
persons and entities to exercise reasonable care to avoid causing
harm to others.” (Maj. opn., ante, at p. 2.) This duty forms the
basis for the general negligence rule that persons and entities
are generally “ ‘ “liable for injuries caused by [their] failure to
exercise reasonable care. (Cabral v. Ralphs Grocery Co. (2011)
51 Cal.4th 764, 771 (Cabral).) “[T]he law imposes a general duty
of care on a defendant only when it is the defendant who has
‘ “ created a risk” ’ ” of harm to the plaintiff, including situations
in which “ ‘ “the defendant is responsible for making the
plaintiff’s position worse.” ’ ” (Brown v. USA Taekwondo (2021)
11 Cal.5th 204, 214 (USA Taekwondo).)
Gilead has argued that it should be exempted from this
standard. In evaluating such a request, courts ask whether the
party can justify the judicial creation of a carveout from the
default duty of care. (Cabral, supra, 51 Cal.4th at p. 783.) In
Rowland v. Christian (1968) 69 Cal.2d 108 (Rowland), this court
affirmed the default duty of reasonable care that is applicable to
everyone. (Id. at p. 112.) It also “identified several
considerations that, when balanced together, may justify a
departure from the fundamental principle embodied in Civil
Code section 1714.” (Cabral, at p. 771.) The Rowland analysis
4
GILEAD TENOFOVIR CASES
Evans, J., dissenting
“is conducted ‘at a relatively broad level of factual generality.’ ”
(Kuciemba v. Victory Woodworks, Inc. (2023) 14 Cal.5th 993,
1021 (Kuciemba).)
The Rowland factors fall into two categories focusing on
foreseeability and public policy concerns. “Three factors —
foreseeability, certainty, and the connection between the
plaintiff and the defendant — address the foreseeability of the
relevant injury, while the other four — moral blame, preventing
future harm, burden, and availability of insurance — take into
account public policy concerns that might support excluding
certain kinds of plaintiffs or injuries from relief.” (Kesner v.
Superior Court (2016) 1 Cal.5th 1132, 1145 (Kesner).) Issues
related to foreseeability are assessed on the basis of information
available at the time of the alleged negligence, while “ ‘our duty
analysis is forward-looking’ in regard to policy issues
surrounding burdens that would be placed on defendants.”
(Kuciemba, supra, 14 Cal.5th at p. 1022.)
The majority concludes that when a drug manufacturer
allegedly has information indicating that a drug is equally
effective and safer than a drug already on the market but elects
to delay development of the alternative to maximize profit, an
exception to the standard duty of care should exist. Such an
exception is not justified.
A. The existence of a duty does not equate to a
finding of liability.
We analyze the Rowland factors to determine “ ‘not
whether they support an exception to the general duty of
reasonable care on the facts of the particular case before us, but
whether carving out an entire category of cases from that
general duty rule is justified by clear considerations of policy.’ ”
(Kuciemba, supra, 14 Cal.5th at p. 1021; see id. at p. 1018 [The
5
GILEAD TENOFOVIR CASES
Evans, J., dissenting
default duty of reasonable care remains the “starting point of
our duty analysis.”].)
We have previously emphasized the clear distinction
between duty and liability. Because our task is to examine
whether a categorical no-duty rule is appropriate, we must take
a general view, rather than one bounded by the facts of the case.
(See Cabral, supra, 51 Cal. 4th at p. 772.) Adhering to this
approach is sometimes challenging because of the overlap
between the Rowland factors and the relevant negligence
elements. For example, “[w]hile the court deciding duty
assesses the foreseeability of injury from ‘the category of
negligent conduct at issue,’ if the defendant did owe the plaintiff
a duty of ordinary care the jury ‘may consider the likelihood or
foreseeability of injury in determining whether, in fact, the
particular defendant’s conduct was negligent in the first place.’ ”
(Id. at p. 773.) Given the potential to conflate the existence of a
duty with a possible finding of liability, courts must be vigilant
in applying the correct frame of analysis. Both Gilead and the
majority fail to do so.
First, Gilead’s request and much of the majority’s
reasoning rely on conjecture about calamitous liability for
Gilead and other pharmaceutical manufacturers. These
arguments misunderstand the fundamental issue. Rejecting an
exception to the default duty does not signal anything about the
merits of plaintiff’s underlying case. Assuming the existence of
a duty would automatically or even likely result in a finding of
liability is contrary to the reality of tort litigation. For example,
we have recognized for decades that restaurants can be liable
for the negligent preparation of food. (See Mexicali Rose v.
Superior Court (1992) 1 Cal.4th 617, 631.) Affirming this
statutory default has not resulted in repeated or ruinous
6
GILEAD TENOFOVIR CASES
Evans, J., dissenting
liability for commercial restaurants. (See id. at p. 632 [noting
this rule “corresponds to modern developments in tort law,”
particularly “our modern emphasis” on section 1714’s
requirement that “all persons . . . use ordinary care to prevent
injuries as the result of their conduct.”].) As discussed below,
pharmaceutical manufacturers have been subject to the
negligence standard of care in other jurisdictions. Yet, Gilead
has not shown that manufacturers have been held to unfair
standards nor has it presented any evidence that societally
beneficial activity has ceased.
Second, Gilead and the majority contradict the procedural
posture of this case, as well as the categorical approach required
by Rowland, by contesting the issue of Gilead’s actual
knowledge of TAF’s safety and efficacy. The majority concludes
without a factual finding on this issue that a drug manufacturer
cannot have actual knowledge of the safety and efficacy of an
alternative before phase III testing. (E.g., maj. opn., ante, at pp.
17–19, 24.) By inappropriately deciding a key factual question
at this phase of the litigation, the majority compounds its errors
in analyzing foreseeability and public policy, as discussed below.
The question of knowledge is further misplaced when one
considers the breadth of the immunity granted by the majority.
The majority’s Rowland analysis appears to foreclose potential
liability even after phase III testing, when actual knowledge is,
in their view, possible.
As required by Rowland, the question here is merely
whether an exception to the default duty of care is warranted
where a pharmaceutical manufacturer has actual knowledge of
a safer and equally effective alternative drug but chooses to
delay the development of that alternative in order to maximize
profit. The existence of a default duty of reasonable care does
7
GILEAD TENOFOVIR CASES
Evans, J., dissenting
not, as the majority fears, equate to a finding of negligence
liability on these facts.
B. Gilead cannot clearly define the scope of its
proposed exception.
Gilead argues that pharmaceutical manufacturers cannot
know when a duty of care “arises” in the development process.
It therefore concludes that the Court of Appeal erred in
imposing an “unbounded” duty to develop safer alternatives. As
an initial matter, this argument misunderstands that section
1714 imposes a universal duty to act reasonably to avoid harm.
The lack of clarity around when a duty would “arise” weighs
against Gilead’s proposed exception. “ ‘No-duty rules are
appropriate only when a court can promulgate relatively clear,
categorical, bright-line rules of law applicable to a general class
of cases.’ ” (Kesner, supra, 1 Cal.5th at p. 1144, quoting Rest.3d
Torts, Liability for Physical and Emotional Harm, § 7, com. a, p.
78.) The absence of any “clear limiting principle” (maj. opn.,
ante, at p. 20) is unnecessary where a default rule exists.
Gilead still argues that a knowledge limitation is
necessary to bound a pharmaceutical company’s duty to its
consumers when making development decisions. Gilead,
however, does not show that a drug manufacturer could never
have actual knowledge that an alternative drug is equally as
effective as a currently available one after phase II testing
(much less after phase III testing). Gilead merely relies on
untested evidence presented by amici curiae and inferences
from the federal regulatory framework. (See maj. opn., ante, at
pp. 17–19.) Because, as the majority opinion notes,
pharmaceutical development is a complex and specialized field,
a definitive pronouncement that knowledge is impossible at this
stage is inappropriate. Gilead’s proposed “narrow” exception
8
GILEAD TENOFOVIR CASES
Evans, J., dissenting
therefore does not create a general class based on relevant
markers like knowledge or culpability. Instead, Gilead
proposes, and the majority accepts, an amorphous carveout for
essentially all development decisions no matter how
unreasonable they may be.
This is despite the fact that plaintiffs’ claim is based on a
duty that exists in the latter stages of development, once a drug
has been “invented” and the manufacturer has done sufficient
testing to gain (assumedly) actual knowledge of its effectiveness.
The Court of Appeal therefore correctly focused on the later
stages of development. (See Gilead, supra, 98 Cal.App.5th at p.
921, fn. 3 [“We use the term ‘invent’ here, rather than ‘develop,’
because the meaning of ‘develop’ in the pharmaceutical context
is ambiguous. Gilead refers to the entire process of drug
creation, from invention through FDA approval, as drug
development. Because plaintiffs’ claim is focused only on the
latter stages of this process, Gilead’s general reference to a ‘duty
to develop’ obscures the precise nature of plaintiffs’ claim.”].)
“While we agree with Gilead that a duty that placed
manufacturers ‘under an endless obligation to pursue ever-better
new products or improvements to existing products’ would be
unworkable and unwarranted,” plaintiffs are not pursuing
liability under such a broad duty. (Id. at p. 921.) The majority,
nonetheless, appears to have foreclosed further development of
any standards in favor of total immunity from negligence
related to drug development.
Plaintiffs do not contend that Gilead was obligated to
develop TAF in its entirety — instead, they argue Gilead acted
unreasonably when it paused development after promising
early-phase results to protect their monopoly and maximize
profits. Therefore, the issue of knowledge is not as dispositive
9
GILEAD TENOFOVIR CASES
Evans, J., dissenting
as Gilead believes. I agree with the Court of Appeal that the
mere possibility of knowledge is a question better left to a fact
finder, and should not be a basis for the creation of a categorical
no-duty rule.
C. The default duty applies across other
complicated contexts and industries.
Because the duty to minimize harm and the accompanying
reasonableness standard apply across a wide range of factual
scenarios, many just as or more complicated than drug
development, Gilead’s and the majority’s concerns about
hindsight bias do not justify an exception. California law
routinely entrusts fact finders with reasonableness
determinations in specialized or causally complex settings
involving intricate technical evidence, multiple parties,
regulatory bodies, and long timelines, using expert testimony
and tailored instructions where needed. For example, in
complex medical malpractice cases, even where the standard of
care is “peculiarly within the knowledge of experts” the question
remains one of fact. (Landeros v. Flood (1976) 17 Cal.3d 399,
410; see also ibid. [“[N]either this nor any other court possesses
the specialized knowledge necessary to resolve the issue as a
matter of law.”].) We have rejected the argument that uncertain
or conflicting scientific evidence justifies a departure from the
reasonable care standard. (Tarasoff v. Regents of University of
California (1976) 17 Cal.3d 425, 437–438; see id. at p. 438
[holding that although therapists argued they were unable to
accurately predict whether patients will act violently, they must
still “exercise ‘that reasonable degree of skill, knowledge, and
care ordinarily possessed and exercised by members of [that
professional specialty] under similar circumstances’ ”].)
10
GILEAD TENOFOVIR CASES
Evans, J., dissenting
Similarly, we have affirmed that in asbestos litigation,
even with long latency periods, multiple products, and scientific
uncertainties, traditional substantial-factor causation is a
question best left to a jury. (Rutherford v. Owens-Illinois, Inc.
(1997) 16 Cal.4th 953, 975–978.) And, in multiparty
construction and workplace accident cases, juries routinely
review blueprints, safety logs and reports, sift through decades
of internal corporate documentation, evaluate contradictory
expert testimony, review state and federal regulations and
inspection reports, and determine complex theories of causation.
(See, e.g., Elsner v. Uveges (2004) 34 Cal.4th 915, 923
[discussing the admissibility of California Occupational Safety
and Health Act of 1973’s (Lab. Code, § 6300 et seq.) provisions
in evaluating negligence actions].) We trust them to do so rather
than simply providing broad immunity from liability.
Our previous decisions applying Rowland confirm that
factual complexity does not support a carveout from the default
duty of care. In Kesner, supra, 1 Cal.5th at page 1140, we
affirmed that employers and landowners must exercise
reasonable care in preventing secondary exposure to asbestos.
(See also id. at p. 1147 [“[D]efendants cite no authority requiring
a scientific consensus to establish foreseeability in the context
of duty analysis.”].) Similarly, in T.H. v. Novartis
Pharmaceuticals Corp. (2017) 4 Cal.5th 145 (Novartis), we
reaffirmed that fact finders are in the best position to evaluate
the reasonableness of a pharmaceutical company’s conduct in
creating warning labels on its products. (See id. at p. 192.)
Drug manufacturing is no different. While
pharmaceuticals are heavily regulated public goods, they are
not so complicated as to be beyond a reasonableness standard.
Gilead and the majority assert that the pharmaceutical industry
11
GILEAD TENOFOVIR CASES
Evans, J., dissenting
is exceptional in its development process, requiring “complex
strategic decisions about where to devote finite resources based
on limited information.” (Maj. opn., ante, at p. 40.) Gilead and
the majority, however, recognize that a drug manufacturer can
be liable in defect. Defect claims are incredibly complex,
requiring fact finders to sift through potentially contradictory
scientific evidence and to evaluate iterative decisionmaking.
Yet California law still empowers juries to make these
determinations. Decisions related to drug development are no
more complicated than those related to whether or not a drug is
defective. The analysis in both contexts requires the
“retrospective evaluations of reasonableness” the majority here
eschews as completely unworkable. (Ibid.)
It is unclear why the majority suddenly doubts the ability
of fact finders to make reasonableness determinations in this
context. (E.g., maj. opn., ante, at p. 41 [“This retrospective
inquiry creates a substantial risk that hindsight will supplant
the manufacturer’s contemporaneous judgment, even when its
decisions were reasonable when made.”].) By this same logic, all
complicated business decisions should be freed from the
hindsight risk of fact finders. Despite the risk of hindsight bias,
California law, however, has long allocated these questions of
liability to juries, not to courts.
III. ROWLAND FACTORS
Contrary to the majority’s conclusion, the Rowland factors
weigh against the creation of a carveout from the default duty
of reasonable care. The harm from intentionally and
unreasonably delaying the release of an equally effective but
safer drug is foreseeable, and public policy supports preserving
the default protections against negligent conduct.
12
GILEAD TENOFOVIR CASES
Evans, J., dissenting
Rowland reinforced “that the basic policy of this state set
forth by the Legislature in section 1714 of the Civil Code is that
everyone is responsible for an injury caused to another” by their
conduct. (Rowland, supra, 69 Cal.2d at pp. 118–119.) The
majority claims “it makes little sense to premise an entirely new
form of liability” on limited precedent. (Maj. opn., ante, at p. 46.)
This framing, however, subverts not only Rowland, but also
California negligence law, as codified by the Legislature. (See
Rowland, at p. 112 [“[I]t is clear that in the absence of [a]
statutory provision declaring an exception to the fundamental
principle enunciated by section 1714 of the Civil Code, no such
exception should be made unless clearly supported by public
policy.”].)
Unlike the majority, I would not foreclose an entire
category of negligence cases based on conjectural claims of
harm, nor on plaintiff’s “failure” to point to other cases where
pharmaceutical manufacturers were found liable for the exact
conduct alleged in plaintiff’s complaint. Recognizing the default
duty of care does not depend on previously proven violations of
that duty. Instead, we should look to basic negligence
principles.
We have never before held that drug manufacturers are
entitled to an exception to the default negligence duty of care.
The majority today applies Rowland with undue deference to
the pharmaceutical industry to the peril of vulnerable and
captive patients. I fear its broad reasoning, while purporting to
be cabined by the facts of this case, risks diluting the
“traditional” negligence claims it avers are left untouched. (Maj.
opn., ante, at p. 55.)
13
GILEAD TENOFOVIR CASES
Evans, J., dissenting
A. The harm to plaintiffs is foreseeable.
Rowland’s foreseeability factors do not support exempting
Gilead from the default duty to act reasonably. Because Gilead
has a monopoly over the research, development, and sale of
tenofovir-based medications, its conduct was necessarily the
root cause of plaintiffs’ alleged harm. Despite Gilead’s attempts
to contest knowledge and offer hypothetical alternative causes,
Gilead’s decision to delay the development of TAF was the first
step in plaintiffs’ continued exposure to harmful side effects. In
a different factual scenario, a pharmaceutical manufacturer
might show that the FDA denied approval or some other
roadblock arose. Gilead, however, asks — and the majority
agrees — to exempt Gilead from any scrutiny related to the
reasonableness of its actions. Gilead claims it is functionally
powerless to control its own development cycle, and therefore
blameless when it allegedly prioritizes profit over preventing
harm. This argument contradicts both logic and our precedents.
“Regardless of alternative sources of [harm], or variations
in the personal precautions [plaintiffs] undertake, it is plainly
foreseeable” that negligent conduct can lead to harm.
(Kuciemba, supra, 14 Cal.5th at p. 1025.) “An intervening third
party’s actions that are ‘themselves derivative of defendants’
allegedly negligent conduct . . . do not diminish the closeness of
the connection between defendant’s conduct and plaintiff’s
injury for purposes of determining the existence of a duty of
care.’ ” (Kesner, supra, 1 Cal.5th at p. 1148, quoting Beacon
Residential Community Assn. v. Skidmore, Owings & Merrill
LLP (2014) 59 Cal.4th 568, 583.)
Particularly relevant in the foreseeability analysis is the
element of control by the would-be defendant over the source of
harm. (Novartis, supra, 4 Cal.5th at p. 168.) In Novartis, we
14
GILEAD TENOFOVIR CASES
Evans, J., dissenting
declined to create an exception from the typical duty to warn
because federal regulations granted the brand-name drug
manufacturer exclusive control over the active ingredients in
the generic drug and the content of the warnings included in the
generic’s label. (Ibid.) So too here, Gilead had complete control
not only over the development and release of TAF, but also all
other tenofovir-based medications. The “intervening conduct” is
“predictable and derivative of the alleged misconduct,” thus
weighing against an exception. (Kesner, supra, 1 Cal.5th at p.
1149.)
Gilead argues that it could not have actual knowledge of
TAF’s safety and efficacy until full FDA approval, thus making
plaintiff’s harm too attenuated. As discussed above, at this
stage of the litigation, a court should not substitute its decision-
making for that of a better-informed fact finder on the issue of
knowledge, and whether such knowledge is possible. Further, it
is unclear why FDA approval would provide a drug
manufacturer with additional knowledge of safety and efficacy,
given that the manufacturer is the one furnishing the FDA with
the safety and efficacy data the agency considers. (See maj.
opn., ante, at p. 19; 21 U.S.C. § 355(d).)
The very purpose of Rowland is to determine whether a
modest carveout to the general duty of care is justified.
(Rowland, supra, 69 Cal.2d at p. 112.) The majority argues that
reserving the question of knowledge for a fact finder would
render the foreseeability inquiry moot. (Maj. opn., ante, at pp.
25–26.) While the question of foreseeability is a matter of law,
we have previously declined to make such a determination
based on contested facts. (See, Kesner, supra, 1 Cal.5th at p.
1147.) Rowland encourages judicial modesty in making these
determinations. I would not, as the majority chooses, decide
15
GILEAD TENOFOVIR CASES
Evans, J., dissenting
that drug manufacturers as a class can never know that an
alternative drug is safer than an existing one at the stage of
research and development at issue, much less at a later one.
The majority further justifies its decision to provide broad
immunity to Gilead by speculating that any harm to plaintiffs
might be exacerbated by causes other than Gilead. (Maj. opn.,
ante, at pp. 23–26.) To be sure, plaintiffs’ category of alleged
injury — continued, unnecessary exposure to devastating side
effects — could be worsened by other events or by the conduct or
choices of third parties. But, again, such determinations should
be addressed on the facts at the liability stage, rather than
through the creation of a categorical no-duty rule. This is
particularly true given Rowland’s purpose of identifying only
narrow exceptions where warranted. (See Kuciemba, supra, 14
Cal.5th at p. 1021 [Rowland asks “ ‘not whether [the factors]
support an exception to the general duty of reasonable care on
the facts of the particular case before us, but whether carving
out an entire category of cases from that general duty rule is
justified by clear considerations of policy.’ ”].) “We have no
occasion to address other arguments defendants might make to
defeat liability. It must be remembered that a finding of duty is
not a finding of liability.” (Kesner, supra, 1 Cal.5th at p. 1157.)
Gilead should be free to raise challenges to causation at a later
stage, but these conjectural, alternative causes of harm do not
warrant a categorical exception to the default duty of care.
B. The category of conduct is morally
blameworthy.
Contrary to the majority’s conclusion, Rowland’s public
policy factors do not support exempting Gilead from the default
duty to act reasonably. Plaintiffs allege that Gilead alone, by
pausing TAF development, caused them harm. At a categorical
16
GILEAD TENOFOVIR CASES
Evans, J., dissenting
level, unreasonably delaying the development of a safer and
equally effective alternative drug is morally blameworthy. We
have regularly recognized that “[t]he overall policy of preventing
future harm is ordinarily served, in tort law, by imposing the
costs of negligent conduct upon those responsible.” (Cabral,
supra, 51 Cal.4th at p. 781.) As in Novartis, a drug
manufacturer with exclusive rights to develop the safer
alternative formulation of a drug is not only in the best position
to take action, but also is the only entity with the ability to
prevent plaintiffs’ alleged unnecessarily prolonged exposure to
these side effects. (See Novartis, supra, 4 Cal.5th at pp. 168–
169.)
Gilead argues that “ ‘the moral blame that attends
ordinary negligence is generally not sufficient to tip the balance
of the Rowland factors in favor of liability.’ ” (Maj. opn., ante, at
p. 33, quoting Adams v. City of Fremont (1998) 68 Cal.App.4th
243, 270, disapproved of on another ground by USA Taekwondo,
supra, 11 Cal.5th 204.) The allegedly unreasonable conduct
here, however, goes beyond ordinary negligence.
Because drug manufacturers financially benefit from the
significant power imbalance between themselves and TDF
patients when prioritizing or delaying certain drugs for
development, this factor weighs against recognizing an
exception. “[M]oral blame is typically found when the defendant
reaps a financial benefit from the risks it has created.
(Kuciemba, supra, 14 Cal.5th at p. 1025.) As discussed, Gilead
has made over $100 billion from tenofovir-based medications,
profits allegedly increased by the intentionally delayed release
of TAF in 2015, two years before its patent on TDF expired.
Further, “[w]e have previously assigned moral blame, and
we have relied in part on that blame in finding a duty, in
17
GILEAD TENOFOVIR CASES
Evans, J., dissenting
instances where the plaintiffs are particularly powerless or
unsophisticated compared to the defendants or where the
defendants exercised greater control over the risks at issue.”
(Kesner, supra, 1 Cal.5th at p. 1151.) The majority rejects
plaintiffs’ showing on these points, instead relying again on its
factual conclusion that Gilead could not have had actual
knowledge of TAF’s safety and efficacy. Then, paradoxically, the
majority claims a categorical approach, noting that the precise
motivations attributed to the defendant do not weigh on the
moral blameworthiness of conduct. (Maj. opn., ante, at p. 34.)
This argument is contrary to the holdings of Kuciemba and
Kesner, where blameworthiness was attributed, at least in part,
based on the reaping of a profit and inherent power imbalance
between plaintiffs and defendants.
It is true that some manufacturers, including Gilead,
could have justifiable, even laudatory reasons for delaying the
development of an alternative treatment. However, specific
challenges to liability are not relevant when a party seeks a
categorical rule. (Kesner, supra, 1 Cal.5th at p. 1157.)
Defendants like Gilead have considerable profit incentives to
extend the lives of their patents. At the same time, drug
manufacturers have a unique form of control stemming from
patent protections over entire compounds. Potential plaintiffs
are particularly vulnerable where, as here, they need the
lifesaving tenofovir compound, available from only one
manufacturer. Defendants should be able to demonstrate the
reasonableness, neutrality, or societal benefits of delay, if they
exist, during later stages of litigation. The majority’s categorical
exemption, however, does not allow any negligence case against
drug manufacturers to reach the merits, no matter how
18
GILEAD TENOFOVIR CASES
Evans, J., dissenting
unreasonably the manufacturer may have acted or how much
harm they may have caused.
The majority is concerned that considering profit and
power, as required by our precedents, would weigh against an
exception to the duty of care in “virtually every negligence action
brought against a business.” (Maj. opn., ante, at p. 35.) This is
a feature of the Rowland analysis, not a bug. By framing the
duty of reasonable care as a universal default duty, Rowland
moved away from status-based privilege to prioritize protecting
people from preventable injuries. (See Rowland, supra, 69
Cal.2d at pp. 118–119 [“[T]he basic policy of this state set forth
by the Legislature in section 1714 of the Civil Code is that
everyone is responsible for an injury caused to another” by their
conduct.].) Thus, it is unsurprising that allegedly negligent
conduct by a large, profitable company when compared to
captive consumers with little power would be found morally
blameworthy. This is not a “ ‘heads I win, tails you lose’ ” (maj.
opn., ante, at p. 54) exercise as the majority alleges but rather a
recognition of Rowland’s shift towards making negligence
claims more accessible.
Gilead and the majority make much of the ameliorative
steps a manufacturer would need to take to bring a product like
TAF to market. (Maj. opn., ante, at pp. 35–36.) However,
plaintiffs are not claiming that Gilead was required to release a
safer alternative to TDF. Rather, they claim Gilead
intentionally and unreasonably delayed the development of a
safer, more effective alternative drug that they had already
created. If a manufacturer were to reach a roadblock, such as
contradictory results in a later trial or failure to secure FDA
approval, they would not be required to go above and beyond the
19
GILEAD TENOFOVIR CASES
Evans, J., dissenting
ordinary standard of care. Instead, manufacturers need only
reasonable justifications for pausing development.
Blameworthiness under Rowland arises because of the
profit motive and power imbalance inherent in a pharmaceutical
monopoly. Gilead “created the risk of harm to plaintiffs by
selling TDF, a drug with harmful side effects.” (Gilead, supra,
98 Cal.App.5th at p. 935.) And Gilead made “ ‘ “plaintiffs’
position worse” ’ ” (USA Taekwondo, supra, 11 Cal.5th at p. 214)
by allegedly deliberately and unreasonably pausing for almost a
decade the development of an alternative drug without the same
adverse effects, in order to reap substantial profits.
C. The societal costs of the default duty of care
are outweighed by the benefits of ensuring
reasonable conduct.
Gilead requests that all pharmaceutical manufacturers be
immunized from negligence liability stemming from their
development decisions, with no regard to consumer harm.
Again, we are meant to analyze the Rowland factors to
determine “ ‘not whether they support an exception to the
general duty of reasonable care on the facts of the particular
case before us, but whether carving out an entire category of
cases from that general duty rule is justified by clear
considerations of policy.’ ” (Kuciemba, supra, 14 Cal.5th at p.
1021, italics added.) Contrary to the majority’s conclusion,
Gilead does not meet this standard, and the default duty of care
should therefore stand.
These factors require that we weigh necessarily uncertain
policy considerations. (E.g., Kuciemba, supra, 14 Cal.5th at p.
1022 [“ ‘[O]ur duty analysis is forward-looking’ in regard to
policy issues.”].) What the majority calls a “vain exercise” (maj.
opn., ante, at p. 55) is required to hold Gilead to the proper
20
GILEAD TENOFOVIR CASES
Evans, J., dissenting
burden of persuasion. (See Cabral, supra, 51 Cal.4th at p. 771
[“[C]ourts should create [an exception] only where ‘clearly
supported by public policy.’ ”].)
Both this opinion and the majority’s make assumptions
about how parties will behave in the future. The difference is I
do not believe speculation, much of which we have previously
rejected, tips the Rowland factors in favor of additional
immunity for the pharmaceutical industry. (Novartis, supra, 4
Cal.5th at p. 173 [“We are equally unpersuaded
by Novartis’s contention that warning label liability would stifle
innovation by substantially raising drug costs and chilling the
development and marketing of new drugs.”].) The Rowland
factors impose the same standard on any party seeking a
categorical no-duty rule. That standard requires more than the
specter of widespread liability. (E.g., Novartis, at pp. 173–175
[rejecting the argument that the default negligence duty would
expose brand-name manufacturers to limitless liability because
the ordinary limiting principles on negligence claims remain].)
I would follow our precedent and find that the costs, both to the
public and to potential defendants, are outweighed or otherwise
too hypothetical to support an exception to the default duty of
care.
This conclusion reflects the reality of the pharmaceutical
industry. Gilead and other manufacturers are incentivized to
maximize the lifespan of their patents to extend their monopoly
over brand-name sales. (Hickey & Ward, Congressional
Research Service, The Role of Patents and Regulatory
Exclusivities in Drug Pricing (Jan. 30, 2024) (Hickey & Ward)
<https://www.congress.gov/crs-product/R46679> [as of Aug. 3,
2026] [“Because the exclusivity that IP rights provide may
enable the rights holder (e.g., a brand-name drug manufacturer)
21
GILEAD TENOFOVIR CASES
Evans, J., dissenting
to charge higher-than-competitive prices for a period of time,
rights holders may have an incentive to lengthen that time
period as much as possible.”]; all Internet citations in this
opinion are archived by year, docket number and case name
at <https://courts.ca.gov/opinions/cited-supreme-court-
opinions>.) The majority, however, accepts without question
that “drug manufacturers already have ‘ample incentive[s] to
release safer products into the marketplace,’ including the
opportunity to gain market share and avoid reputational harm
associated with adverse effects.” (Maj. opn., ante, at p. 37.) It
therefore concludes that this existing profit-motive structure
undermines the need to “impose” a tort duty to ensure
reasonable decisionmaking about whether and when to develop
safer alternative drugs. (Ibid.)
Decades of litigation and regulation, particularly in the
antitrust context, contradict this assumption that drug makers
are adequately incentivized to release safe and effective drugs
as quickly as possible. (See, e.g., FTC v. Actavis, Inc., (2013) 570
U.S. 136, 140 [Federal Trade Commission action targeting
reverse payment settlements, under which a patentee agrees to
pay a potential generic manufacturer not to produce a version of
the drug until the patent’s term expires].) Where, as here, the
manufacturer has an exclusive patent on the existing product
and its related formulations, negligence liability serves as one
of the few safeguards against unreasonable conduct.
Keeping in mind the level of generality we must utilize in
the Rowland test, I believe the question is whether there is a
societal benefit or cost to maintaining the default duty when a
drug manufacturer has the exclusive patent on a drug and its
similar formulations. Because this monopoly posture means the
traditional incentives do not apply here, I would find the default
22
GILEAD TENOFOVIR CASES
Evans, J., dissenting
duty of reasonable care is a necessary safeguard against
harmful conduct — a purpose it was designed to serve.
Because of the societal benefits of drug development, we
have in the past excepted pharmaceutical manufacturers from
certain kinds of liability. Gilead fails, however, to justify an
additional and especially broad exception, particularly in light
of these pre-existing carveouts. In Brown v. Superior Court
(1988) 44 Cal.3d 1049 (Brown), we exempted prescription drug
manufacturers from strict liability based on the societal cost of
such a duty. (Id. at p. 1063.) We reasoned that the risk of strict
liability could discourage research and development of beneficial
pharmaceuticals or force companies to conduct additional,
unnecessary testing before releasing a drug to market. (Ibid.)
By contrast, a duty here would not disincentivize the
development of drugs but merely allow for the possibility of
liability when the development of less harmful alternatives is
unreasonably delayed in order to maximize profit. Imposing
strict liability, which significantly lowers a plaintiff’s burden,
presents a heightened chance of finding actual fault — a concern
not found here, where plaintiffs still must prove negligence. The
application of the default duty of care would simply encourage
companies to take consumer harm into reasonable account when
making development decisions.
Additionally, the risk of disincentivizing the development
of new drugs or drug formulations is too hypothetical to support
a categorical exemption from liability. Drug makers have long
claimed that any risk of liability would discourage innovation or
shutter entire companies. Gilead, however, overstates the
empirical support for this claim, as long-term studies of the
economic effects of tort liability for pharmaceuticals present
contradictory evidence. (See Garber, Economic Effects of
23
GILEAD TENOFOVIR CASES
Evans, J., dissenting
Product Liability and Other Litigation Involving the Safety and
Effectiveness of Pharmaceuticals (RAND Institute for Civil
Justice, 2013) p. 62 (RAND Study) [“Both proponents and
opponents of preemption have cogent arguments and/or
empirical evidence to support at least some of their claims about
economics effects. Thus, it appears that there are both
substantial social benefits and substantial social costs of . . .
litigation. We cannot, however, compare the social benefits and
costs quantitatively.”].) Such conjecture cannot support an
exception.
For example, Gilead argues that instead of researching
new cures for perhaps rarer or less-studied conditions, it will be
forced to develop alternatives to existing treatments. (See, e.g.,
maj. opn., ante, at p. 43 [“[T]he duty may encourage
manufacturers to focus on making marginal improvements to
existing drugs rather than pursuing novel therapies for diseases
that lack effective treatment.”].) In its discussion on treatments
for rarer conditions, the RAND Study concluded that “it seems
likely that even if product liability deters some product-
development efforts, product liability exposure is unlikely to tip
the balance in many cases.” (RAND Study, supra, at p. 55.)
Because fairly small markets are a “defining characteristic” of
these types of drugs, “it seems likely that there would typically
be insufficient financial incentives to induce companies to invest
in drug development even if there were no liability exposure.”
(Ibid.)
Some evidence, moreover, supports positive societal
effects from tort liability. Research has suggested that “the
potential for litigation to uncover inappropriate corporate
behavior increases the potential costs of engaging in such
behavior and thus, at least in theory, will tend to deter
24
GILEAD TENOFOVIR CASES
Evans, J., dissenting
manufacturers from engaging in such practices.” (RAND Study,
supra, at p. 60; see also ibid. [discussing one study that
examined “examples of alleged practices — such as failing to
report adverse events to the FDA, downplaying the extent or
severity of side effects, failing to make clinical trial data public,
and delaying public reports of adverse events”].) Despite such
mixed empirical evidence, the majority accepts wholesale Gilead
and amici curiae’s unsupported claim that innovation will be
stifled by the mere risk of liability for negligence.
The majority’s acquiescence is further surprising
considering that we recently declined to create a categorical no-
duty rule based on nearly identical speculative “burdens” on
pharmaceutical innovation. (Novartis, supra, 4 Cal.5th at p.
173.) We noted that the pharmaceutical company in that matter
offered “ ‘no clear or sufficient basis for concluding that research
and development will inevitably decrease’ ” and no evidence that
drug innovation had declined in the decades that drug
manufacturers were exposed to strict liability for failure to
warn. (Ibid.) Accordingly, we held that this factor did not
support an exception to the default duty of care. (Ibid.) The
majority attempts to distinguish Novartis by claiming that
manufacturer conduct related to drug development is different
than manufacturer conduct related to already available drugs.
While the contexts may differ, it’s unclear why we should
countenance unreasonable, harmful conduct in one context but
not the other, given California’s strong public policy of
protecting its residents from preventable injuries. (Kuciemba,
supra, 14 Cal.5th at p. 1021 [requiring “ ‘clear considerations of
policy’ ” to justify “ ‘carving out an entire category of cases from
that general duty rule.’ ”].)
25
GILEAD TENOFOVIR CASES
Evans, J., dissenting
In any case, drug makers would need only to avoid
unreasonable decisions regarding whether and when to
commercialize a safer drug. Allocating resources to address a
more pressing medical need instead of developing a safer
alternative, for example, would almost certainly defeat a claim
of negligence. As discussed above, Gilead and the majority,
however, erroneously equate duty with liability. (See Cabral,
supra, 51 Cal.4th at pp. 772–773.)
D. Pharmaceuticals are entirely unlike other
consumer industries.
I find similarly unconvincing Gilead’s and the majority’s
attempted analogies to other industries, including automobile
manufacturing. Rejection of a carveout here has little, if any,
relevance for other products or industries. For example, unlike
drug makers, automobile manufacturers are not given a state-
guaranteed monopoly on producing cars, nor on species of safety
technology. Drivers are free to choose between different
manufacturers and models. However, many drug consumers —
including the HIV patients seeking relief here — have no choice
but to purchase the brand-name medication because patent
protections ensure no other option exists. As a result, drug
manufacturers can charge higher-than-competitive prices for
their products. This dynamic creates a unique incentive for
pharmaceutical companies to extend their periods of exclusive
sales. Drug companies have developed well-known tactics to
extend these rights for as long as possible.2
2
“Such patenting practices include so-called (1) patent
‘evergreening,’ (2) ‘product hopping,’ (3) ‘patent thickets,’ and (4)
‘pay-for-delay’ settlements. Patent ‘evergreening’ is the alleged
practice of filing for new patents on secondary features of a
26
GILEAD TENOFOVIR CASES
Evans, J., dissenting
While regulators have sometimes stepped in to punish or
curtail such conduct, they have done so mostly in individual
enforcement actions targeting antitrust violations. Of course,
anticompetitive or anti-consumer behavior in the industry at
large does not automatically entitle plaintiffs to relief. That
said, it is an additional reason why a categorical exemption for
the pharmaceutical industry is both unwarranted and unwise.
Apart from industry-wide distinctions in regulation, drugs
themselves are unique. Plaintiffs here needed TDF to prevent
HIV from reproducing and eventually contributing to their
deaths. They could not decline the side effects, which are part
and parcel of the life-saving drug. By contrast, users of products
like cars can pay more for added safety features or abstain from
purchasing a car altogether. Because of these product-specific
considerations, leaving the default duty of reasonable care in
place here is unlikely to have negative effects on other markets.
E. The burden on pharmaceutical
manufacturers is negligible.
Drug manufacturers, denied an exception to the default
duty of reasonable care, will not lose the profit incentive to
pharmaceutical as earlier patents expire, thereby extending
effective patent exclusivity past the original 20-year term.
‘Product hopping’ is the alleged practice of a brand
manufacturer attempting to switch the market to a new, similar
product covered by later-expiring patents before IP rights on an
existing product expire. ‘Patent thickets’ refer to portfolios of
numerous, overlapping patents on the same pharmaceutical,
which allegedly deter competition due to the risk of
infringement and the high cost of patent litigation. ‘Pay-for-
delay’ or ‘reverse payment’ settlements resolve patent litigation
through payments or other compensation from a brand to a
generic or biosimilar manufacturer to delay generic market
entry.” (Hickey & Ward, supra, summary.)
27
GILEAD TENOFOVIR CASES
Evans, J., dissenting
develop better versions of their products, nor to develop novel
therapies, if they are simply required to make reasonable
decisions to minimize harm. They merely would “lose” the
ability to time those decisions without regards to injury to
consumers. This factor, accordingly, does not support an
exception. The majority, on the other hand, concludes that
maintaining the default duty to act reasonably would force
manufacturers to devote significant time, money, and resources
to developing and commercializing drugs they might not
otherwise prioritize. (Maj. opn., ante, at pp. 43–45.)
Gilead fails to show that the default duty of care imposes
any actual burden. Gilead could have acted exactly as it did here
and be protected from liability if it behaved reasonably. Drug
manufacturers theoretically have been subject to this default
duty of care for decades and yet drug development (including the
parallel testing of “backup” formulations) has not ground to a
halt.
The opioid crisis and resulting litigation shed light on the
insufficiency of Gilead’s arguments. While opioid plaintiffs
brought suit under several different theories of liability, in at
least one case, the negligence claims brought by the State of
Rhode Island against a group of opioid manufacturers survived
summary judgment. The court there held that “[d]efendants
owed duties to the State, including ‘a duty to exercise reasonable
care in manufacturing, marketing, selling, and distributing
highly dangerous opioid drugs,’ as well as ‘a duty to exercise
reasonable care . . . not to cause foreseeable harm to others.’ ”
(State v. Purdue Pharma L.P. (R.I., Aug. 16, 2019, No. PC-2018-
4555) 2019 WL 3991963, p. *16.) This acknowledgement of the
default duty highlights the societal benefits of flexible tort
liability. The opioid crisis has caused incalculable harms to
28
GILEAD TENOFOVIR CASES
Evans, J., dissenting
individuals, organizations, and public entities. The combination
of litigation by private and public parties with governmental
regulations started the long process of addressing these injuries
and preventing future harm. (See also Ohio County Commission
v. Express Scripts, Inc. (N.D.W.Va., Dec. 23, 2024, No. 5:24-CV-
142) 2024 WL 5701504, p. *15 [“[P]laintiffs have identified that
defendants owed duties to plaintiffs, including ‘a duty to employ
reasonable standards of care in the sale, delivery, dispensing,
promotion, and gatekeeping control of the supply of highly
addictive, dangerous opioids. This duty includes a duty to not
create a foreseeable risk of harm or injury.’ ”].) Like opioid
plaintiffs, the consumers here seek redress under the default
general duty of care for defendant’s alleged conduct in the
development of tenofovir products. Given the role negligence
liability has played in addressing the opioid epidemic, it is
disappointing that the majority has provided such sweeping
immunity to drug manufacturers.
Even assuming there would be a financial cost to
defendants, such a cost likely would be small in comparison to
profits as well as expected as part of their research and
development process.3 The burden here is importantly distinct
from cases like Verdugo v. Target Corp. (2014) 59 Cal.4th 312.
In contrast to training retail employees on a particular medical
intervention (id. at p. 340), pharmaceutical companies already
anticipate the costs of developing, researching, and testing new
drugs. Since Gilead eventually paid the costs to conduct a phase
3
Plaintiffs submit evidence that at the end of 2003 Gilead
had cash and marketable securities (effectively cash) of over
$700 million. Gilead’s internal documents indicated that the
remaining costs for approval of TAF were estimated to be $82
million.
29
GILEAD TENOFOVIR CASES
Evans, J., dissenting
III trial, seek FDA approval, and market TAF, it is disingenuous
to argue that the cost of further developing an alternative drug
is an onerous burden that would otherwise not be incurred.
Moreover, drug manufacturers directly benefit from the
investment in research and development, as demonstrated by
the more than $100 billion dollars Gilead has made from TDF
and TAF-based drugs. Therefore, Gilead’s alleged financial
burden does not support a categorical no-duty rule.
For the same reasons, Gilead fails to show that difficulty
accessing insurance supports an exception. (E.g., maj. opn.,
ante, at p. 52 [“[T]he duty would likely result in significantly
greater liability exposure than the Court of Appeal perceived.”].)
The sky has not yet fallen on drug manufacturers, despite years
of potential exposure to negligence liability for development
decisions. Neither the majority nor Gilead offers “ ‘clear
considerations of policy’ ” which justify “ ‘carving out an entire
category of cases from that general duty rule.’ ” (Kuciemba,
supra, 14 Cal.5th at p. 1021.)
IV. ALTERNATIVE AVENUES
The majority believes that its decision will not entirely
insulate manufacturers from potential liability because
plaintiffs have alternative avenues to seek redress. (Maj. opn.,
ante, at pp. 55–56.) Negligence liability, however, is plaintiffs’
clearest avenue to compensation for their alleged injuries.
Plaintiffs cannot adequately protect themselves via statutes
prohibiting unlawful, unfair, fraudulent, or misleading conduct.
They are seeking accountability for a company’s alleged
unreasonable prioritization of profit, conduct which purportedly
harmed its captive consumer base of TDF users. They request
compensation for their individual injuries allegedly caused by
30
GILEAD TENOFOVIR CASES
Evans, J., dissenting
many years of unnecessary exposure to devastating side effects.
Their claim sounds clearly in negligence.
Although I understand my colleagues’ wariness of
potentially “modifying” incentives in such a complicated and
crucial industry, I suggest that recognizing an exemption from
the typical tort scheme is the far more disruptive outcome. The
Pennsylvania Supreme Court, in declining to immunize
pharmaceutical companies from negligence claims, noted that
“Congress and the Legislature are in the best position to make
these sorts of weighty and consequence-laden policymaking
judgments impacting a traditional, state-law, civil, remedial
scheme.” (Lance v. Wyeth (2014) 624 Pa. 231, 277 [85 A.3d 434].)
In general, I tend to agree that regulation by other
branches of government may be preferrable to civil liability.
Regulatory measures, such as FDA review and approval of new
drugs, however, cannot address the harm of manufacturers
withholding safer alternatives.4 (See maj. opn., ante, at p. 55.)
Even assuming an overlap, public and private regulation can
and should coexist, especially in the pharmaceutical context.
State tort actions play a pivotal role “ ‘as a complementary form
of drug regulation’ with respect to drug labeling.” (Novartis,
supra, 4 Cal.5th at p. 169, quoting Wyeth v. Levine (2009) 555
U.S. 555, 578; see Wyeth, at p. 579 [“State tort suits uncover
unknown drug hazards and provide incentives for drug
4
While drug safety oversight has been a core function of the
FDA, recent changes in leadership, staffing, budget priorities,
and policy direction have raised concerns about the continuing
vitality and effectiveness of the regulatory framework. (See,
e.g., Gottlieb, Making the FDA Great Again (June 18, 2026)
JAMA Health Forum, vol. 7, No. 6 <https://jamanetwork.com/
journals/jama-health-forum/fullarticle/2850771> [as of Aug. 3,
2026].)
31
GILEAD TENOFOVIR CASES
Evans, J., dissenting
manufacturers to disclose safety risks promptly. They also
serve a distinct compensatory function that may
motivate injured persons to come forward with information.”].)
Regulations alone may not adequately protect consumers from
harm. (See Novartis, at p. 169, citing Stevens v. Parke, Davis &
Co. (1973) 9 Cal.3d 51, 65.) Accordingly, plaintiffs should be
able to plead their claim in negligence, as we have regularly
confirmed that pharmaceutical companies are subject to such
liability. (E.g., Brown, supra, 44 Cal.3d at p. 1069, fn. 12
[Manufacturers can and should still be liable “under general
principles of negligence.”].)
V. CONCLUSION
Today’s decision provides sweeping immunity, allowing
pharmaceutical companies to develop drugs without accounting
for the risk of harm to consumers like plaintiffs here, who are
held captive when a drug is both lifesaving and subject to
exclusive manufacturing rights.
I urge the Legislature to consider whether pharmaceutical
companies should have immunity from negligence liability no
matter how unreasonably they may act or how much serious and
avoidable harm they may cause patients when making decisions
about commercializing drugs.
EVANS, J.
32
See next page for addresses and telephone numbers for counsel who
argued in Supreme Court.
Name of Opinion Gilead Tenofovir Cases
__________________________________________________________
Procedural Posture (see XX below)
Original Appeal
Original Proceeding
Review Granted (published) XX 98 Cal.App.5th 911
Review Granted (unpublished)
Rehearing Granted
__________________________________________________________
Opinion No. S283862
Date Filed: August 3, 2026
__________________________________________________________
Court: Superior
County: San Francisco
Judge: Andrew Y.S. Cheng
__________________________________________________________
Counsel:
Sidley Austin, Debra Pole, Joshua Anderson, Sean Commons, David
Carpenter, Collin Wedel; Orrick, Herrington & Sutcliffe, Elizabeth
Bixby, E. Joshua Rosenkranz, Andrew Silverman, Naomi Scotten,
Siobhan Atkins, Cesar Lopez-Morales and Emily Villano for Petitioner.
Richard A. Epstein for Pacific Research Institute as Amicus Curiae on
behalf of Petitioner.
Blank Rome and Ana Tagvoryan for Atlantic Legal Foundation as
Amicus Curiae on behalf of Petitioner.
Tucker Ellis, Mollie F. Benedict and Peter L. Choate for International
Association of Defense Counsel as Amicus Curiae on behalf of
Petitioner.
Cory L. Andrews and John M. Masslon II for Washington Legal
Foundation as Amicus Curiae on behalf of Petitioner.
DLA Piper, Ben C. Fabens-Lassen, Ilana H. Eisenstein, Alicia Hickock,
M. David Josefovits, Adam Pierson and Justin R. Sarno for the
Chamber of Commerce of the United States of America, the California
Chamber of Commerce, the Washington Legal Foundation and the
National Retail Federation as Amici Curiae on behalf of Petitioner.
Covington & Burling, Ashley M. Simonsen, Alice L. Phillips, Michael
X. Imbroscio, Paul W. Schmidt and Gregory L. Halperin for
Pharmaceutical Research and Manufacturers of America,
Biotechnology Innovation Organization and Advanced Medical
Technology Association and California Life Sciences as Amici Curiae
on behalf of Petitioner.
Buchalter, Mary-Christine Sungaila and Paul A. Alarcon for Product
Liability Advisory Council, Inc., as Amicus Curiae on behalf of
Petitioner.
Gutierrez, Preciado & House and Calvin House for the Civil Justice
Association of California, the California Manufacturers & Technology
Association, the California Business Roundtable, the Bay Area Council
and Biocom California as Amici Curiae on behalf of Petitioner.
Ian Adams for the International Center for Law and Economics as
Amicus Curiae on behalf of Petitioner.
Reed Smith, James C. Martin, Lisa M. Baird, Corinne Fierro, Steven J.
Boranian and Sarah B. Johansen for DRI-Center for Law and Public
Policy, the Association of Defense Counsel of Northern California and
Nevada and the Association of Southern California Defense Counsel as
Amici Curiae on behalf of Petitioner.
Gibson, Dunn & Crutcher, Theane D. Evangelis and Daniel R. Adler
for the National Association of Manufacturers, the Alliance for
Automotive Innovation, the American Tort Reform Association, the
American Coatings Association, the American Chemistry Council,
Medical Devices Manufacturers Association and the Consumer
Technology Association as Amici Curiae on behalf of Petitioner.
Paul, Weiss, Rifkind, Wharton & Garrison, Randall S. Luskey, Anna
M. Stapleton, Jackson Willis, Kannon K. Shanmugam; and Colin L.
Ward for Viasat, Inc., Textron Inc., Uber Technologies, Inc., VIZIO,
Inc., and Lyft, Inc., as Amici Curiae on behalf of Petitioner.
Bowman and Brooke and Paul A. Alarcon for Product Liability
Advisory Counsel, Inc., as Amicus Curiae on behalf of Petitioner.
O’Melveny & Myers, Charles C. Lifland, Sabrina H. Strong, Jason
Zarrow, Jeffrey L. Fisher; Cohen Williams and Marc S. Williams for
Archer Aviation, Inc., Bayer U.S. LLC, Becton, Dickinson and
Company, Biogen Inc., Bristol Myers Squibb Company, Corteva
Agriscience LLC, Cytokinetics, Incorporated, The Dow Chemical
Company, DuPont de Nemours, Inc., Eli Lilly and Company, GE
Healthcare Technologies, Inc., Genentech Inc., General Motors LLC,
Glaukos Corporation, GSK LLC, Hamilton Beach Brands, Inc.,
Hyundai Motor America, Incyte Corporation, Johnson & Johnson, Inc.,
Kenvue Inc., Kia America, Inc., Organon & Co., Medtronic, Inc., Merck
& Co, Inc., Pfizer, Inc., Regeneron Pharmaceuticals, Inc., Roche
Molecular Systems, Inc., Sanofi US, Sonoma Biotherapeutics, Inc.,
STORM Therapeutics Ltd., Takeda Pharmaceuticals U.S.A., Inc.,
Toyota Motor North America, Inc., Vertex Pharmaceuticals Inc.,
Volkswagen Group of America, Inc., and Zimmer Biomet Holdings,
Inc., as Amici Curiae on behalf of Petitioner.
Quinn Emanuel Urquhart & Sullivan and John Potter for Community
Education Group, C. Virginia Fields, Global Coalition on Aging, HIV
and Hepatitis Policy Institute, Liver Coalition of San Diego, Dr.
Eugene McCray, National Minority Quality Forum, Partnership to
Fight Chronic Disease and Phill Wilson as Amici Curiae on behalf of
Petitioner.
No appearance for Respondent.
Berger Montague, F. Paul Bland, Joseph E. Samuel, Jr., Julie A.
Pollock and Matthew I. Summers for Justice Catalyst Law as Amicus
Curiae on behalf of Respondent and Real Parties in Interest.
Grant & Eisenhoffer, M. Elizabeth Graham; Jenner Law, Robert K.
Jenner; Moskovitz Appellate Team, Myron Moskovitz; Kershaw, Cook
& Talley, Kershaw Talley Barlow, William A. Kershaw; Schneider
Wallace Cottrell Konecky, Amy Eskin; Esner, Chang, Boyer & Murphy,
Andrew N. Chang, Holly N. Boyer and Kiran R. Iyer for Real Parties in
Interest.
The Arkin Law Firm, Sharon J. Arkin; and Jeffrey R. White for the
American Association for Justice and Consumer Attorneys of
California as Amici Curiae on behalf of Real Parties in Interest.
Karla Gilbride for Public Justice as Amicus Curiae on behalf of Real
Parties in Interest.
Thomas A. Myers and Jonathan M. Eisenberg for AIDS Healthcare
Foundation as Amicus Curiae on behalf of Real Parties in Interest.
Emory Law School Supreme Court Advocacy Program, Paul Koster
and Kay Levine for Professors of Public Health and Bioethics as Amici
Curiae on behalf of Real Parties in Interest.
Counsel who argued in Supreme Court (not intended for
publication with opinion):
E. Joshua Rosenkranz
Orrick, Herrington & Sutcliffe LLP
51 West 52d Street
New York, NY 10019
(212) 506-5380
Holly N. Boyer
Esner, Chang, Boyer & Murphy
600 South Lake Avenue, Suite 408
Pasadena, CA 91106
(626) 535-9860