No Negligence Duty to Accelerate Development/Commercialization of an Unapproved “Safer” Drug When the Marketed Drug Is Concededly Nondefective

1. Introduction

Gilead Tenofovir Cases arises from allegations concerning two antiretroviral drugs used in HIV treatment: tenofovir disoproxil fumarate (TDF) and tenofovir alafenamide fumarate (TAF), both developed by Gilead Sciences, Inc. Plaintiffs alleged they suffered serious renal, bone, or tooth injuries from taking TDF, but they conceded TDF is not defective and did not contend it should have been withdrawn from the market.

Instead, plaintiffs advanced a novel negligence theory: Gilead allegedly “unreasonably delayed” taking the steps necessary to bring TAF to market—an alternative drug plaintiffs claimed was “equally effective and less toxic”—and plaintiffs alleged they would have switched to TAF and avoided injury had it been available sooner.

The core issue was thus not the safety of the marketed product under traditional products-liability defect concepts, but whether California negligence law can impose liability for injuries caused by a concededly nondefective drug based on the manufacturer’s allegedly unreasonable research-and-commercialization timeline for a different drug that was not yet approved.

2. Summary of the Opinion

Procedural posture

After the trial court denied summary judgment on negligence (and fraudulent concealment), Gilead sought writ relief. The Court of Appeal (in Gilead Tenofovir Cases (2024) 98 Cal.App.5th 911) ordered summary adjudication on fraudulent concealment but allowed the negligence claim to proceed, reasoning a manufacturer’s duty of reasonable care can extend beyond not marketing defective products and that the Rowland factors did not justify an exception on the record presented.

Disposition

The Supreme Court of California reversed and directed entry of summary judgment for Gilead on all causes of action.

Key holding

The court declined to definitively decide whether negligence liability can ever attach for injuries from a nondefective product. But even assuming such a duty might exist in some contexts, the court held the Rowland v. Christian factors require an exception here, because:

• Before large-scale human trials and FDA approval, it is not reasonably foreseeable—based on information available at the time—that delaying commercialization of a candidate drug will harm users of an existing drug.
• The connection between any delay and injury is attenuated by uncertain scientific outcomes and discretionary third-party decisions (later trial outcomes, FDA approval, physicians’ individualized prescribing decisions).
• Imposing a duty would create substantial burdens and adverse community consequences, including chilling innovation and distorting R&D priorities.

Separate writings

• Guerrero, C. J., concurring in the result: The negligence claim fails because plaintiffs did not allege a defect; under Jiminez v. Sears, Roebuck & Co. and Merrill v. Navegar, Inc., defect is essential.
• Kruger, J., concurring: Agrees precedent generally requires defect, but joins the majority’s assumption arguendo and its Rowland exception analysis.
• Evans, J., dissenting: Characterizes the decision as “sweeping immunity,” would apply Civil Code section 1714’s default duty and allow negligence claims when a manufacturer delays commercialization of a known safer, equally effective alternative for profit.

3. Analysis

3.1 Precedents Cited

A. Civil Code section 1714 baseline duty, and judicial refinement of “reasonable care”

The Court of Appeal relied on section 1714’s general duty of care. The Supreme Court agreed section 1714 supplies a baseline duty, but emphasized that reasonable care is often defined and limited by “statutes, regulations, or judicial decisions,” citing Parsons v. Crown Disposal Co. (1997) 15 Cal.4th 456 and Ramirez v. Plough, Inc. (1993) 6 Cal.4th 539. This move matters: the majority framed products-liability doctrine as a body of judicial decisions that already cabins what “reasonable care” means for manufacturers with respect to product-caused harm.

B. Products liability’s defect “linchpin”

The majority’s skepticism about plaintiffs’ theory drew from defect-centered products-liability authorities:

• Trejo v. Johnson & Johnson (2017) 13 Cal.App.5th 110 (manufacturer duty: design, manufacture, and market products free from defects).
• Merrill v. Navegar, Inc. (2001) 26 Cal.4th 465 (defect requirement applies in both negligence and strict products liability; defect is essential to recover).
• Barker v. Lull Engineering Co. (1978) 20 Cal.3d 413 (risk-benefit and consumer-expectations tests for design defect in strict liability, generally for non-prescription-drug products).
• Jiminez v. Sears, Roebuck & Co. (1971) 4 Cal.3d 379 (negligent design/manufacture claims still require that negligence resulted in a defective product).
• Brown v. Superior Court (1988) 44 Cal.3d 1049 (no strict liability for prescription drug design defects; negligence remains, but within defect-based framework; policy concern about chilling innovation).

Chief Justice Guerrero’s concurrence made these precedents dispositive, quoting the “black letter” statement from Merrill v. Navegar, Inc. (2001) 26 Cal.4th 465 that “a plaintiff must prove that a defect caused injury,” and invoking Jiminez v. Sears, Roebuck & Co. (1971) 4 Cal.3d 379 to emphasize defect as an “essential element” of negligence against a manufacturer.

C. Rowland duty-exception framework and categorical analysis

The doctrinal engine of the majority opinion is Rowland v. Christian (1968) 69 Cal.2d 108, which allows exceptions to section 1714’s baseline duty based on foreseeability and policy. The majority emphasized modern “categorical” duty analysis, relying on Kuciemba v. Victory Woodworks, Inc. (2023) 14 Cal.5th 993 for the proposition that the duty inquiry is categorical rather than case-specific.

The court also used other duty/foreseeability authorities to stress remoteness and intervening actor concerns:

• Cabral v. Ralphs Grocery Co. (2011) 51 Cal.4th 764 (foreseeability at category level; connection factor akin to proximate cause).
• Kesner v. Superior Court (2016) 1 Cal.5th 1132 (foreseeability is key; duty analysis at broad generality).
• Vasilenko v. Grace Family Church (2017) 3 Cal.5th 1077 (attenuated connection where injury depends on independent third-party actions).
• State Dept. of State Hospitals v. Superior Court (2015) 61 Cal.4th 339 (speculative chain of discretionary decisions undermines proximate connection).
• Elden v. Sheldon (1988) 46 Cal.3d 267 (policy may bar a claim “no matter how foreseeable the risk”).

D. The Court of Appeal’s reliance on a non-defect negligence example rejected as inapt

The Court of Appeal had leaned on Mexicali Rose v. Superior Court (1992) 1 Cal.4th 617 to argue negligence can exist absent a defect. The Supreme Court distinguished Mexicali Rose as a “unique context of food preparation” and noted it was expressly limited to “commercial restaurant establishments.” The majority treated this limitation as undermining any attempt to generalize a non-defect negligence theory to pharmaceutical R&D decisions.

E. Prescription-drug information duties and hindsight concerns

In addressing plaintiffs’ “we would have switched” assertions, the court analogized to contexts where hindsight bias is a recognized danger:

• Himes v. Somatics, LLC (2024) 16 Cal.5th 209 (rejecting subjective patient standard in failure-to-warn due to hindsight bias).
• Cobbs v. Grant (1972) 8 Cal.3d 229 (informed consent; subjective patient standard rejected due to hindsight bias).

3.2 Legal Reasoning

A. The court’s central move: assume arguendo, then negate duty via Rowland

Rather than squarely holding “no defect, no negligence” (as the Chief Justice would), the majority chose a narrower—yet still categorical—route:

1. Acknowledge strong products-liability precedent tying manufacturer duty to defects and express “significant doubts” about plaintiffs’ theory.
2. Assume arguendo a broader section 1714 duty could exist.
3. Apply Rowland to recognize an exception for claims seeking to impose negligence liability for delayed development/commercialization of an unapproved alternative drug where the marketed drug is concededly nondefective.

This structure allows the court to decide the case without formally rewriting the broader doctrinal debate about whether defect is always required, while still foreclosing this particular “negligent failure to timely innovate/commercialize” theory in the drug context.

B. Foreseeability: knowledge limits in early-stage development

The majority grounded its foreseeability analysis in the FDA regulatory pathway (IND; phases I–III; NDA) and the empirical reality of high attrition. The doctrinal point is not that injuries from TDF are unforeseeable in general, but that the relevant alleged negligence—delaying development of TAF at an early stage—cannot be treated as foreseeably causing injury because:

• Early trials (phase I/II) are preliminary; long-term or rare adverse effects may not appear until phase III.
• FDA approval requires “substantial evidence,” generally at least two adequate and well-controlled studies (often phase III), and manufacturers cannot legally market investigational drugs as safe/effective.
• Thus, as a class, manufacturers cannot “know” at early stages that a candidate is safer and equally effective in a definitive comparative sense, undermining foreseeability “at the time of the alleged negligence” (per Kuciemba v. Victory Woodworks, Inc.).

C. Connection/proximate-cause analog: attenuated chain with intervening discretionary decisions

The majority treated the “closeness of connection” factor as strongly favoring no duty because harm depends on a chain of contingencies and discretionary acts beyond the manufacturer’s control:

• Subsequent trial success or failure (scientific uncertainty).
• FDA approval (independent regulatory judgment).
• Physician-patient prescribing decisions and individualized risk-benefit tradeoffs (echoing Brown v. Superior Court on patient-specific “superiority”).

This is the doctrinal bridge from negligence theory to administrability: the longer and more discretionary the chain, the less suitable the category is for imposing a duty in the first place.

D. Public policy: innovation, resource allocation, and the “second-guessing” concern

The opinion treats the pharmaceutical context as uniquely vulnerable to harmful effects of a broad negligence duty over R&D timelines:

• Chilling effects: manufacturers may avoid studying backup candidates to reduce litigation risk.
• Distorted incentives: prioritization might skew toward incremental improvements to existing therapies (to mitigate duty exposure) at the expense of novel therapies.
• Perverse delay incentives: fear of duty-driven liability could lead to withholding otherwise nondefective, lifesaving drugs until “perfect” alternatives are fully explored—an echo of Brown v. Superior Court’s warning about delayed release pending additional science.
• Extraordinary burden: the alleged duty would effectively compel costly phase III trials and commercialization investments for candidate drugs based on early signals, despite high failure rates and uncertain benefit.

E. Moral blame: categorical neutrality and the “reasonable ameliorative steps” concept

The majority resisted moral-blame analysis tied to plaintiffs’ allegations about profit motive, emphasizing that duty analysis is categorical. It also stressed a pragmatic point from cases like Parsons v. Crown Disposal Co. and Vasilenko v. Grace Family Church: moral blame typically increases when there are feasible and effective “ameliorative steps” within the defendant’s control. Here, the purported ameliorative step—accelerating development—requires years, immense cost, uncertain outcomes, and is partly contingent on FDA action.

F. The concurrences and dissent illuminate the doctrinal fork

The separate opinions expose two competing ways to conceptualize the case:

• Guerrero (result-only): This is straightforward products-liability negligence: no defect alleged, so no claim. The concurrence criticizes the majority for creating “unnecessary uncertainty” by assuming a broader duty and then crafting an exception.
• Kruger (concurrence): Agrees precedent strongly suggests defect is required, but accepts the majority’s assumption arguendo while endorsing the administrability and policy concerns.
• Evans (dissent): Treats the case as classic negligence: section 1714 default duty applies; juries can handle complexity; profit-plus-monopoly plus avoidable harm supports moral blame and future-harm prevention. The dissent frames the majority as granting “sweeping immunity” and urges legislative intervention.

3.3 Impact

A. Immediate doctrinal effect in California

The decision squarely forecloses—at least in the circumstances described—a negligence theory premised on delay in developing/commercializing an unapproved allegedly safer alternative drug when the injury-causing drug is conceded nondefective. Plaintiffs will be pushed toward established avenues:

• Negligent failure to warn / inadequate labeling (e.g., under principles cited in Carlin v. Superior Court (1996) 13 Cal.4th 1104 and Stevens v. Parke, Davis & Co. (1973) 9 Cal.3d 51).
• Manufacturing defect or warning defect theories.
• Negligent design within the constraints applicable to prescription drugs (and the limits recognized in Brown v. Superior Court).
• Consumer protection or fraud-based claims where properly pled and supported (though here fraudulent concealment was eliminated below).

B. “Negligent failure to innovate” and “tort regulation of R&D timelines” largely rejected

The ruling is best understood as rejecting tort supervision of pharmaceutical R&D sequencing when the theory is untethered to defect in the marketed product and depends on early-stage comparisons, uncertain science, and regulatory discretion.

C. Spillover considerations (signaled, not decided)

The majority expressly warned that a broad duty theory could have far-reaching consequences across industries (e.g., automotive safety technology), although it did not decide duties outside pharmaceuticals. The opinion will nonetheless be cited by defendants in other innovation-heavy sectors as authority against “delay-to-improve” negligence theories.

D. Internal tension left in place: defect requirement vs Rowland exception

Because the majority assumed arguendo rather than held defect is always required, future litigants may continue to test whether non-defect negligence theories exist in other contexts. Yet, the Chief Justice’s concurrence supplies a ready-made road map for defense arguments that defect remains an essential element under Merrill v. Navegar, Inc. and Jiminez v. Sears, Roebuck & Co..

4. Complex Concepts Simplified

Civil Code section 1714 (baseline duty of reasonable care)

Section 1714 states the general principle that everyone must use ordinary care to avoid injuring others. Courts start from that “default” duty, but may recognize categorical limits (exceptions) when policy and foreseeability considerations justify them.

The Rowland factors (when courts carve out “no duty” categories)

Under Rowland v. Christian, courts consider foreseeability-related factors (foreseeability, certainty of injury, closeness of connection) and policy factors (moral blame, preventing future harm, burden to defendant/community, insurance). Importantly, modern California duty analysis is “categorical,” meaning courts ask whether the whole class of cases should be subject to duty, not whether the defendant acted badly in this specific case.

Products liability “defect” (why it matters)

In products cases, “defect” functions as a gatekeeping concept: liability generally turns on whether the product was defectively designed, defectively manufactured, or accompanied by inadequate warnings. Plaintiffs here conceded the product they took (TDF) was not defective, so they attempted to shift the case from product defect to “unreasonable delay” in offering a different product.

FDA drug development phases (why the court focused on them)

• Phase I: small group; early safety/tolerability/dosing.
• Phase II: more patients; early efficacy and short-term side effects.
• Phase III: large, longer studies; most robust safety/efficacy data.
• FDA approval (NDA): required before marketing a new drug.

The court used this framework to explain why early signals do not reliably establish that a drug “is safer and equally effective,” and why harm from delayed commercialization is not sufficiently foreseeable at the time of early-stage development decisions.

Intervening decisions and “attenuated connection”

Even if an alternative drug might someday be safer, whether a particular plaintiff would have avoided injury depends on many contingencies: later trial outcomes, FDA approval, and individualized physician/patient choices. When too many independent, discretionary steps separate the defendant’s conduct from the plaintiff’s injury, courts often treat the connection as too remote to impose a duty for the category of conduct.

5. Conclusion

Gilead Tenofovir Cases establishes a significant limit on negligence theories that seek to convert pharmaceutical R&D and commercialization timing into tort duties—at least where plaintiffs concede the marketed drug is nondefective and the alleged safer alternative had not completed large-scale testing or received FDA approval. The Supreme Court, applying Rowland v. Christian categorically, held that imposing such a duty would rest on uncertain science, attenuated causal chains, and policy burdens that risk distorting innovation and public health outcomes.

The concurrences and dissent underscore an important doctrinal tension: whether the case should have been resolved on the simpler ground that defect is an essential element (Merrill v. Navegar, Inc.; Jiminez v. Sears, Roebuck & Co.) or, as the majority did, by assuming a broader duty and then crafting a Rowland-based exception. Either way, the practical takeaway is clear: California will not recognize negligence liability for delaying an unapproved “safer” alternative drug when the drug actually used is conceded nondefective.