Photo of Stephen McConnell

We do not quite fit the stereotype of the fat cat – at least (we hope), not anymore.  But there was a time when at least one evening per week would be marked by scotch and cigars. Our antiquity and iffy physical constitution have now reduced such festivities to two or three times per annum. (We missed last week’s Bobby Burns Night Dinner; missed the poetry, missed the peaty firewater, and missed the conviviality of old friends, some of whom, sadly, have passed on.) Still, as the most interesting man in the world might say, while we do not always smoke cigars, when we do, we like Rocky Patel cigars. So imagine our curiosity when Bexis sent us a case with Rocky Patel in the caption. The case is Rocky Patel Premium Cigars, Inc., et al. v. Bonta, 2025 WL 3903972 (S.D. Cal. Dec. 23, 2025).  That “et al.” on the plaintiff side includes some other fine cigar makers, such as A. Fuente, as well as distributors and trade associations. Fuente cigars, such as the Hemingway Short Story, are splendid. We had a rooting interest in this case. The defendant, Bonta, is the California Attorney General. So we really had a rooting interest. On the theory of smoke’em if you’ve got’em, we could not resist scribbling about the Rocky Patel case this week, even if its application to our daily practice is even more iffy than our geezerish health.

We previously blogged about a tobacco decision that applied Buckman preemption against an Iowa consumer fraud law clamping down on vape tobacco products. The Iowa law created a registration regime that explicitly incorporated federal premarket standards as a determinative factor for market access. That is, Iowa was asserting enforcement authority that belonged exclusively to the Food and Drug Administration (FDA). Hello, Buckman preemption. Iowa’s law would have had a better chance of surviving if it was a flat-out ban.

Now we come to today’s case, a decision from the Southern District of California that goes the other way, but still offers backhanded support for Buckman preemption in the drug/device area.  The Rocky Patel case concerns California’s ban of flavored tobacco products.  The statute at issue is Cal. Health & Safety Code section 104559.5. The California AG established and maintained a list of unflavored tobacco products (“UTL”).  Any tobacco product not appearing on the UTL after it is published shall be deemed a flavored product banned by the statute. There is an out “if the FDA has indicated that the product does not require such approval.”

The cigar sellers argued that “premium cigars, as defined by federal regulations, are exempt from the federal premarket tobacco product application process” and, therefore, they cannot be flavored products within the meaning of the California statute. Just a little thought should lead one to the conclusion that premium cigars are not the sort of flavored tobacco products that the California solons had in mind. It is not as if kids are looking to puff on $30 cigars because they have the flavor of .. of what?  Not fruit or vanilla or treacle; rather, it is the flavor of tobacco. Really good tobacco. It is a grown-up flavor. It is an acquired taste. And, just possibly, it is a flavor preference that is none of the government’s business. People gripe about Big Tobacco. What about Big Nanny? What would Winston Churchill and Sigmund Freud say about California’s silly law? (Or have they both been canceled?)

In any event, the cigar sellers challenged the California statute.  They asserted five causes of action: (1) violation of the First and Fourteenth Amendments; (2) violation of the dormant Commerce Clause; (3) violation of the Supremacy Clause (implied preemption); and (5) violation of the Due Process Clause.  The complaint sought injunctive relief. The first test for a preliminary injunction is likelihood of success. In assessing that likelihood, the court focused on express preemption, implied preemption, and the First Amendment.

The federal law at issue is the Family Smoking Prevention and Tobacco Control Act (the TCA), which amended the Food, Drug, and Cosmetic Act (FDCA) to give the FDA regulatory authority over aspects of tobacco. The structure of the TCA makes preemption something of an adventure. The TCA contains a broad “preservation” clause allowing states to regulate more stringently (including outright bans) than federal law, followed by multiple preemptive carveouts, in turn followed by a savings clause allowing state regulation of tobacco “sales.” This byzantine preemption/savings structure renders the express preemption aspect of the Rocky Patel opinion not perfectly analogous to our drug/device sandbox. 

First, the court held that the plaintiff’s express preemption argument was unlikely to succeed. The issue was whether the state’s UTL statute and implementing regulations “establish requirements relating to premarket review that are different from or in addition to those established by the TCA … or whether the UTL regime establishes requirements relating to a permissible sales ban on flavored products, such that it falls within the Preservation and Savings Clauses.” This issue turned on an interpretation of what “premarket review” means. The court ultimately decided that the preservation of “premarket review” as a basis for express preemption would cover a fraud on the FDA claim or a state redefinition of categories subject to PMA, but not this flat sales ban. 

As for implied preemption, the court agreed that a state enforcement system “parasitic” on the FDCA (like the Iowa vaping statute in our earlier blogpost) would be preempted.  But the court deemed California’s flavored tobacco ban as being non-parasitic.  The court distinguished (and somewhat demeaned) the Iowa decision. According to the Rocky Patel court, California’s ban on flavored tobacco was a public health measure, and thus the dreaded presumption against implied preemption applied.  As the court saw it, the federal tobacco regulation generally constituted a mere floor that state laws could exceed.  A state ban does not rely on FDA regulatory standards as an enforcement mechanism.  Since the federal tobacco statute intended to preserve a large role for the states, nothing in drug (Nexus Pharmaceuticals) or device (Buckman) cases prohibited state reliance on FDA determinations to enforce its ban.  The ban does not rely on the FDCA; rather, it relies on historic state power to govern tobacco sales and the statutorily established federal regulatory floor for regulation of tobacco products.  In the context of tobacco sales regulation, incorporating federal standards into a state sales’ requirement is not “enforcement” of federal law. 

The plaintiffs argued that the California statute “restricts the manner in which premium cigar manufacturers and importers may describe their products and thus their commercial speech.” The court rejected this First Amendment argument, but at least did acknowledge that the ban triggered First Amendment scrutiny under the Central Hudson test for evaluating restrictions on commercial speech. In applying that test, the Rocky Patel court held that that statute’s “creation of a rebuttable presumption where a manufacturer describes its product as having a characterizing flavor is narrowly tailored to serve the State’s interest in banning flavored tobacco products.” Moreover, the UTL “does not prohibit manufacturers from using any descriptions but merely seeks to further the goal of ferreting out flavored products by drawing special attention to these products that the manufacturer itself describes as being flavored.” 

That “merely” is doing a lot of work.  But we think the “ferreting” word is appropriate. Maybe Big Nanny is also Big Ferret. Maybe the California UTL law is constitutional, but it is also, as Justice Stewart said in his Griswold dissent, an “uncommonly silly” law.   

Photo of Michelle Yeary

Some cases feel brand new. Others feel like a remake you didn’t ask for—but somehow know all the lines to already. This one is firmly in the latter category. For those of us who have been around long enough to remember dial-up internet and the original pedicle screw litigation, this case reads like a throwback. In fact, one of this blogger’s very first cases (working with Bexis) was the Bone Screw MDL—the one that gave us Buckman Co. v. Plaintiffs’ Legal Committee. So, Elowson v. Alphatec Spine Inc., 2026 U.S. Dist. LEXIS 13591 (D. Ariz. Jan. 26, 2026), a negligence per se claim based entirely on alleged FDCA violations, involving pedicle screws, gave us a distinct sense of déjà vu. Spoiler alert: Buckman still wins.

The plaintiff underwent spinal surgery in 2019 to remove a tumor on her lumbar spine. Part of the procedure required spinal fusion using a plate-and-screw system. Things went along just fine—no complaints, no complications, no litigation—for half a decade. Then, in 2024, plaintiff began experiencing pain. An x-ray revealed that one screw had broken at some point before the bone fully fused. Id. at *1.

Cue the lawsuit. Never mind that the device lasted five years. Never mind that the product warnings expressly state that screws may fail if fusion does not occur. Never mind that pedicle screws are subject to the immutable laws of physics. Plaintiff sued anyway, alleging among other things negligence per se based entirely on supposed violations of the FDCA and FDA regulations.

That claim did not survive defendant’s motion to dismiss. And for good reason. Plaintiff’s negligence per se theory was simple in concept and fatal in execution. According to plaintiff, the manufacturer had a duty to update the device design because it was based on an allegedly “outdated” “European style of design” and because U.S. manufacturers supposedly use a design that reduces the risk of fracture. Id. at *4-5. Plaintiff alleged that the “source” of that duty was the FDCA and its regulations, id. at *5-6, rather than any traditional state-law tort principle. That is a problem.  A Buckman-sized problem.

As the court correctly held, negligence per se claims premised exclusively on alleged FDCA or FDA regulatory violations are preempted. Full stop. The Supreme Court settled this decades ago. Id. at *7. If Buckman had a walk-on song, this would be its moment.

The negligence per se claim here would not exist without the FDCA. Plaintiff did not identify any independent state-law duty requiring manufacturers to update designs based on evolving regulatory expectations or comparative international practices. Instead, plaintiff attempted to bootstrap FDA regulations into a state tort claim. That is exactly what Buckman forbids: private litigants bringing private enforcement actions of the FDCA under the guise of state law.

Negligence per se is not a free-floating cause of action. It applies only where a defendant violates a “specific legal requirement, not a general standard of care.” The plaintiff’s problem was that the only “specificity” she could point to came from FDA regulations. Without those regulations, there was no duty, no breach, and no claim.

Courts have consistently recognized that claims not tied to traditional state tort duties—but instead dependent on federal regulatory schemes—are preempted. Allowing them to proceed would invite juries to second-guess FDA regulatory decisions and enforcement priorities. That, again, is Buckman territory. The court declined that invitation.

The court did, however, allow plaintiff’s strict liability design defect claim to survive the pleadings stage. Defendant argued that Arizona follows either the Second or Third Restatement of Torts and that under either it would reject strict liability design defect for a medical device. On the former, the court concluded that whether comment k applies—whether defendant’s screws are unavoidably unsafe—is a question of fact that can’t be resolved on the pleadings. Id. at *11-12. On the latter, the court disagreed with defendant that Restatement Third §6c was Arizona law. Under §6c, a strict liability design defect claim would require plaintiff to plead that a reasonable health-care provider, knowing the risks and benefits of the medical device, would not prescribe the device to any class of patients. Id. at *12. The Arizona Supreme Court, however, in a case that pre-dates the Restatement Third, adopted the standard risk/utility test for strict liability design. As a federal court sitting in diversity, the Elowson court was bound to apply that standard. Id. at *13. Plaintiff’s allegations sufficiently alleged a design defect (the “outdated” European style); an alternative design (the “smooth” U.S. style), and that defendant’s “failure to design screws that remain intact and do not break caused her injuries.” Id. at *15. That was enough to survive a TwIqbal challenge.

But perspective is important. Pleading survival is not victory—it is a timeout. The case now proceeds with discovery, expert scrutiny, and the inconvenient reality that spinal fixation devices, by their very nature, carry a risk of mechanical failure. To date, no bone screws—whether designed in the U.S., the U.K., or on Mars——are immune from breakage, especially in the absence of fusion. Gravity and biomechanics remain undefeated. Warnings here explicitly addressed the precise risk that materialized. The device functioned for five years. And the alternative design theory exists, for now, only on paper. In other words, this looks far more like a temporary setback than a turning point.

The more important takeaway is that this decision reinforces a familiar but critical principle: plaintiffs cannot manufacture negligence per se claims by repackaging alleged FDCA violations. Buckman remains the OG pedicle screw case—and it still has teeth. Sometimes, the old rules still work—much like those screws did, for half a decade.

Photo of Bexis

At the recent ACI Drug and Medical Device annual conference, Bexis created something of a stir by broaching the subject of litigation discovery into the “prompts” that are typically used to create output from generative artificial intelligence.  A fair number of the attendees apparently had not considered that possibility.  Well, it’s already being done, and counsel need to be prepared to address it.

What caselaw exists – so far exclusively from AI copyright litigation − treats prompts as one more form of ediscovery, discoverable (or not) under generally applicable discovery principles.  In Concord Music Group, Inc. v. Anthropic PBC, 2025 WL 1482734 (N.D. Cal. May 23, 2025), for example, the defendant was compelled to “produce a total of 5 million prompt-output pairs” to “be drawn equally from pre-suit and post-suit data” and “randomly selected.”  Id. at 4.  Non-lawyer AI prompts by a party’s employees have also been ordered produced, subject to usual proportionality constraints.  Concord Music Group, Inc. v. Anthropic PBC, 2025 WL 2267950, at *1-2 (N.D. Cal. Aug. 8, 2025) (ordering production of prompt/output generated by the defendant’s “founder, executive or managing agent” and any other “identif[ied]” employee).  Irrelevant AI prompts and their results are, by definition, non-discoverable.  In re OpenAI, Inc., Copyright Infringement Litigation, 800 F. Supp.3d 602, 611-12 (Mag. S.D.N.Y. 2025). Also in the Concord Music litigation, counsel was held to have waived the work product privilege to the extent that they had turned over AI prompts to an expert witness. Concord Music Group, Inc. v. Anthropic PBC, 2025 WL 3677935, at *3 (N.D. Cal. Dec. 18, 2025).

Application of general legal principles also means that, to the extent that counsel are using AI – and prompting it – in the conduct of litigation, the prompts and the resulting output (referred to as “prompt/output” or “prompt/output pairs”) can be protected by discovery by work-product principles.  Concord Music also addressed this issue and rejected an argument that attorney prompts were unprivileged as “unpersuasive.”

[Defendant’s] initial argument, that the information it seeks (undisclosed prompts and outputs, and the settings therefore) is not privileged is unpersuasive.  [Plaintiffs] cite cases where courts . . . have found precisely this information to constitute attorney work product.  [Defendant] distinguishes only [the] denial of waiver, . . . but does not distinguish the basic finding that the failed prompts and related settings are attorney work product.  This Court agrees. . . .

Id. at *2 (citations omitted).  See Tremblay v. OpenAI, Inc., 2024 WL 3748003, at *2 (N.D. Cal. Aug. 8, 2024) (“ChatGPT prompts were queries crafted by counsel and contain counsel’s mental impressions and opinions about how to interrogate ChatGPT, in an effort to vindicate Plaintiffs’ [case]”).

That’s all we’ve found, since this is a quite new area.  We do note, however that the Tenth Circuit recently ordered a litigant to identify AI prompts used in preparing a brief, citing its “inherent power,“ but that was in the context of sanctioning that party for hallucinated citations. Moore v. City of Del City, 2025 WL 3471341, at *2 (10th Cir. Dec. 3, 2025). Non-lawyers – particularly corporate employees, need to appreciate that their AI prompts and the resultant outputs can be discoverable to the same extent as other electronically stored information.  Counsel need to know that they can waive work product protection of prompt/output pairs if they do not treat litigation-related AI use as confidential.  Defense counsel, in particular, need to ensure that our clients are prepared for potential discovery into AI prompts and their results.

Photo of Eric Hudson

We previously posted about plaintiffs’ shenanigans in attempting to defeat diversity in a medical device case removed to the Northern District of Illinois. Plaintiffs’ antics were unsuccessful, and the federal court denied plaintiff’s motion remand.  Today’s decision addresses the defendants’ motion to dismiss plaintiffs’ third amended complaint in the same case, and it is very favorable from a preemption standpoint. Miller v. Rush University Medical Center, 2026 WL 147413 (N.D. Ill. Jan. 20, 2026).

Miller involves a cervical disc replacement device (the “Mobi-C”). The device includes an accompanying inserter with a “depth stop” mechanism used to prevent the Mobi-C from going too far into the spinal cord during insertion. Plaintiff and his spouse sued the device manufacturer, the distributor, two sales representatives, the hospital, and the treating physician based on claims that he suffered spinal cord contusions during surgery. The defendants moved to dismiss, with the manufacturer and distributor arguing that all claims against them were preempted. The court agreed.  

Continue Reading PMA Preemption in the Northern District of Illinois
Photo of Lisa Baird

December 9, 2026?  What is that about? 

Well, December 9, 2026 is the deadline for each member nation within the European Union to have incorporated the EU’s new Product Liability Directive (“PLD”) into their own national laws.  In key markets like Germany, that work is well underway.

Not every EU member state will meet the deadline, and we doubt that courts within the E.U. will instantaneously be overrun with rabid plaintiff-side product liability lawyers from the United States, but the date is significant nonetheless because the EU—for reasons that boggle us—has decided that Europe simply does not have enough product liability litigation, and the EU PLD is the solution they hit upon to address this “problem”.

We have raised the alarm about the new EU Product Liability Directive before, for good reason.  Formally known as Directive (EU) 2024/2853 of the European Parliament and of the Council of 23 October 2024 on liability for defective products and repealing Council Directive 85/374/EEC (Nov. 18, 2024), the EU PLD signifies a seismic shift in product liability exposure in Europe.

The EU PLD establishes a no-fault liability regime for “economic operators”—including manufacturers, importers, authorized representatives, fulfilment service providers, and, in certain cases, distributors and online platforms—for damage caused by defective products.

The definition of “product” also is expansive, and includes not only tangible goods but also software, digital manufacturing files, raw materials, and electricity.  So it is not just companies that are used to product liability exposure in the U.S. that need to pay attention, because the EU PLD treats software as a product for product liability purposes and digital services integrated into or inter-connected with products, such as software updates or health monitoring services, are potentially within the scope of liability if they affect product safety.  Free and open-source software supplied outside commercial activity is excluded, but if such software is integrated into a product in the course of commercial activity, the manufacturer of the final product may be held liable. 

The point being that in addition to manufacturers and sellers of what we think of as products, even tech companies, social media, suppliers of medical devices with software functionality or services, also might want to assess their risk, insurance, and contracts with the EU PLD in mind.

Other key points of the EU PLD include:

Types of Compensable Damage:

Compensation is available for death or personal injury (including medically recognized psychological harm), damage to or destruction of property (excluding the defective product itself and property used exclusively for professional purposes), and destruction or corruption of non-professional data. Pure economic loss, privacy infringements, and discrimination do not trigger liability under this Directive, though other liability regimes may apply.

Determining Defectiveness:

A product is considered defective if it fails to provide the safety that the public is entitled to expect, assessed objectively and taking into account factors such as intended use, foreseeable misuse, presentation, technical features, expected lifespan, and compliance with safety requirements—including cybersecurity.

The EU PLD specifically addresses products with high safety expectations, such as life-sustaining medical devices, and allows courts to find all devices defective just because they come from the same production series as another defective device.

Presumptions and Burden of Proof:

Claimants under the EU PLD technically must prove defectiveness, damage, and causation, but the Directive has several rebuttable presumptions that really lighten the claimants’ evidentiary burdens.

•           Defectiveness is presumed if the defendant fails to disclose relevant evidence, if the product does not comply with mandatory safety requirements, or in cases of “obvious malfunction”.

•           Causation is presumed where the type of damage is typically caused by the defect in question.

•           Courts may presume defectiveness or causation (or both) where technical or scientific complexity makes it hard for the claimant.  Seriously. 

Defendants retain the right to rebut these presumptions, and the first order of business under this new product liability regime will be for defendants to establish good law about how this is done.  Member States must publish final appellate judgments, and the EU Commission will maintain a public database of relevant case law, so good authorities potentially will help across all member states—and bad precedent potentially will hurt the same way.

Liability and Exemptions:

Liability is joint and several where multiple economic operators are involved. Economic operators may be exempted if they prove, among other things, that the defect did not exist when the product was placed on the market, that defectiveness resulted from compliance with legal requirements, or that the state of scientific and technical knowledge at the relevant time could not have revealed the defect (“development risk defence”). However, Member States may elect to forgo the development risk defence for specific product categories in the public interest.

Manufacturers remain liable for defectiveness arising after market placement if it results from software, related services, or lack of necessary updates within their control.

Liability cannot be contractually limited or excluded, and national laws may not set financial ceilings.

Circular Economy and Substantial Modifications:

The Directive addresses circular economy practices, holding those who substantially modify products outside the original manufacturer’s control liable as manufacturers of the modified product. Substantial modification includes changes to performance, purpose, or risk profile not foreseen in the initial risk assessment.

Statutes of Limitations and Repose:

The limitation period for claims is three years from the date the injured person became aware of the damage, defectiveness, and liable operator. An “expiry period” of ten years applies from market placement, extended to 25 years for latent personal injuries.

There is little to like in the EU PLD.  Add in the fact that many EU countries have markedly different approaches to attorney-client privilege and work product confidentiality, changes that make representative or group actions more available, and the interplay between other EU regulatory regimes and the EU PLD, and a storm is brewing. 

We are still only under a watch, not a warning which, in weather forecaster public messaging terms, means the tortillas, meat, salsa and other ingredients are ready, but the taco is not yet on the table. 

Use the time remaining ‘til December 9, 2026 to batten down the hatches.

Photo of Stephen McConnell

How many times have you seen a lawyer end the trial direct examination (or deposition redirect) of his/her expert by perfunctorily asking, “Do you hold all your opinions to a reasonable degree of certainty?” Then there is the obligatory “Yes.” The magic words have been uttered.  All is right with the world, right?

Maybe. 

If the expert has done a good job of describing the opinions and supporting analyses, those magic words might not be necessary. More interestingly, if the supporting analyses are frail, or the opinions are hedged with modifiers (less politely described as weasel-words)  the magic words might not be sufficient.  

For the second week in a row, we are discussing a Pennsylvania Supreme Court decision in a criminal matter.  Last week, it was an affirmance.  This week, in Commonwealth of Pennsylvania v. Fitzpatrick, 2026 WL 157732 (Pa. Jan. 21, 2026), it is a reversal of a murder conviction. The issue is whether an expert’s testimony about the manner (not scientific cause) of death, even though such expert testimony is not required by law to support the prosecution’s theory of manner of death, must satisfy the “reasonable certainty” requirement that applies to expert testimony. The answer is Yes. Thus, Fitzpatrick is yet another criminal case with significant civil implications. 

The defendant in Fitzpatrick was accused of murdering his wife.  She drowned in a tributary of the Susquehanna River after an accident involving an all-terrain vehicle (ATV). Husband and wife were supposedly riding together.  The husband claimed that the ATV flipped backwards, tossing them both into the creek.  The husband managed to climb out of the creek unscathed. The wife was not so lucky 

There was no doubt that drowning was the cause of the wife’s death.  The issue, according to the Supreme Court, was the manner of death – accident, homicide, suicide, etc.  At the beginning of the opinion, we are treated to a long discussion of the difference between cause and manner of death. It feels very scholarly in an old-timey way. And then we noticed that the author was Justice Wecht. He was one of the three Democratic Justices who won a retention election last November.  He is a double Yalie, and that shows in the intellectual heft in the Fitzpatrick opinion. The opinion is well written. Plus, the subject matter of the case is fascinating. It is a good read. Think of the Restatement as rewritten by Mick Herron. In any event, while the prosecution needs an expert to opine as to the cause of death (here, drowning), whether the drowning happened via accident or malevolence is an issue that a jury can decide without help from an expert. 

Nevertheless, the Commonwealth offered that help to the jury, via a prosecution expert witness (an expert on aquatic deaths) who testified using the “magic words” “reasonable degree of medical certainty” to certain opinions concerning the manner of death.  Based on the nature of the wife’s injuries and the husband’s complete absence of injuries, the prosecution expert opined that the wife did not die because the ATV accidentally flipped over but, rather, because the husband administered blunt force injuries and then deposited his wife into the creek. (There was other evidence against the defendant, such as his wife’s written statements that she feared her husband wanted to kill her, the husband’s affair with another woman, the existence of a life insurance policy, the husband’s internet searches on polygraphs and life insurance, and the husband’s lie about the existence of his wife’s cell phone. But that evidence was separate and apart from the expert’s opinions regarding manner of death.)

The problem is that on cross examination, the prosecution expert admitted that his opinions were held to a “more likely than not” standard – a situation often seen in civil litigation.  The expert deployed some very unmagic words such as “possibility,” unlikely,” “I don’t think,” and “consistent with.” The trial court excluded the testimony, but the Superior Court, on appeal, reversed and invented a novel exception to the reasonable degree of medical certainty requirement for manner of death; because such expert testimony was not, strictly speaking, required to establish that element of the crime, the jury could hear such expert testimony if it was “sturdy.”

The Supreme Court reversed.  We get some history as to the origin of the “reasonable certainty” standard.  Those words are absent in Pa. R. Evid. 702. The concept originated in a Chicago personal injury case in 1916.  It did not appear in a Pennsylvania case until 1968.  The standard saw a “virtual explosion of usage” in the 1970’s, to the point where it became “unmistakably clear” that Pennsylvania “case law, evidentiary rules, and supporting secondary materials require that an expert hold his or her opinion to a reasonable degree of certainty in order to be admissible in legal proceedings.”  (It might be entertaining and useful for you to conduct a similar archeological expedition as to the origin of the “reasonable certainty” standard in your jurisdiction.) 

The Fitzpatrick court held that there are no exceptions to the requirement that all expert opinions must be held to a reasonable degree of professional certainty, which in the case of medical testimony means “medical certainty.”   The Supreme Court also definitively held that mere “more likely than not” or “could have” testimony does not rise to the level of reasonable certainty and thus must be excluded.  Further, the entirety of the relevant opinion evidence must be considered, and rote pronouncement of the “magic words” is not controlling.  

The Supreme Court rejected the Superior Court’s invention of the lesser “sturdy” standard in any area of expert testimony.  The term is incapable of being reliably and consistently applied. (That is an excellent and unanswerable point.) It is not enough that the opinion be reached using the generally applicable standards of the relevant profession, it must also be held to a reasonable degree of certainty.  

Therefore, the prosecution expert opinion in Fitzpatrick was inadmissible, even though the expert used the magic words, because of his admissions on cross-examination.  “[M]ore likely than not falls below the level of certainty required for the admission of expert testimony.” 

While few readers of this blog will work on murder cases, Fitzpatrick might be useful in any case where the other side’s expert plays the usual game of using the “magic words” of “reasonable certainty” as a fig leaf covering what is really a mere “more likely than not” opinion. If failure to satisfy the actual “reasonable certainty” standard can overturn a murder case, it should also prevent a private plaintiff from inflicting junk science on a jury. 

Photo of Michelle Yeary

If you ever needed proof that timing is everything, the Taxotere litigation has you covered.

Last month, a court denied summary judgment to the brand manufacturer, finding that it allegedly acquired “newly acquired information” post-dating Taxotere’s original FDA approval in 1996. This month, however, the very same court granted summary judgment to the manufacturers of docetaxel, the active ingredient in Taxotere, whose products were approved in 2011 via §505(b)(2) of the FDCA.   

The key distinction—the clock didn’t start running in 1996 for everyone. In re Taxotere (Docetexal) Eye Injury Products Liability Litigation, 2026 WL 123664 (E.D. La. Jan. 16, 2026).

To start, a §505(b)(2) approval is something of a hybrid between a New Drug Application (brands) and an Abbreviated New Drug Application (generics). It is available for drugs that differ from the RLD (reference listed drug) “in ways that are slight enough for the manufacturer to still rely on the RLD’s safety and efficacy data.” The manufacturer submits an NDA, but it “need contain only that information needed to support the modification(s) of the listed drug.  Unlike [generic] drugs, §505(b)(2) drugs are not required to use the exact same labeling as the RLD.” Id. at *2. So yes we are talking about generics. No, we are not talking about Mensing preemption.

The labels for both Taxotere and docetexal warned of “excessive tearing which may be attributable to lacrimal duct obstruction,” but plaintiffs argue it should have said more. Such claims are preempted, however, unless plaintiffs can point to newly acquired information that would have permitted a unilateral label change (i.e. without the FDA’s approval) under the FDA’s Changes Being Effected (“CBE”) regulation.  In the case of docetexal, plaintiffs argued that information that post-dated the original Taxotere approval but pre-dated the §505(b)(2) approvals should be considered newly acquired. Which if allowed would almost certainly have led the court to reach the same conclusion it did in December.

But here’s where we get a little help from the Fifth Circuit and the first Taxotere MDL. In Hickey v. Hospira, 102 F.4th 748 (5th Cir. 2024), which we talk about here, the appellate court found that “the newly acquired information inquiry centers on what happens after the FDA approves a §505(b)(2) drug.” In re Taxotere, 2026 WL 123664, *9 (emphasis in original). While the §505(b)(2) manufacturers can rely on the brand manufacturer’s safety and efficacy data, they don’t have a “right of reference” to it, meaning they don’t get to see all of it. You can’t say what is “new” if you don’t have a point of reference to compare it to. Id. at *8. That’s why their clock did not start in 1996. Instead, the Fifth Circuit said the proper “baseline comparator” for §505(b)(2) manufacturers is the “publicly available literature that existed prior to the §505(b)(2) manufacturers’ approval.” Id. at *10.

Once the right timing for the baseline was set, the rest of the decision is a standard CBE analysis—did the §505(b)(2) manufacturers have newly acquired information that post-dated the 2011 approval of their drugs that would have supported a unilateral label change. Id.  Without going into the details of the science, the answer was no. Id. at *10-14.

Under the CBE regulation, post-approval information qualifies as “newly acquired” only if it reveals “risks of a different type, or greater severity or frequency” than what was already known and disclosed at the time of approval. Id. at *10. That’s a high bar—and intentionally so. The FDA, not juries, decides what risks warrant warnings, and manufacturers can only change labels unilaterally when genuinely new safety information emerges.

Here, the science tells the same story it did before approval—just with more footnotes. And no, calling old data “new” doesn’t magically make it so. Relying on new articles that “summarize pre-approval literature” is simply repackaging existing knowledge. Id.at *13. If the post-approval literature confirms what was already known, that’s called confirmation, not innovation. And confirmation does not authorize a CBE label change. Because the manufacturers could not have changed their labels without prior FDA approval, plaintiffs’ state-law failure-to-warn claims collapse under impossibility preemption.

While different, the two Taxotere rulings are not contradictory. They’re chronological. And while we think the December ruling was strained on what constitutes newly acquired information, what may count for a product approved in 1996 does not automatically remain “new” forever. By 2011, the FDA had already seen it, digested it, and approved products with that knowledge in hand. When nothing new emerges after approval, manufacturers—brand or generic—cannot be faulted for failing to warn about risks the FDA already considered.

Photo of Steven Boranian

Simply charging a price higher than what plaintiffs want for an effective and non-defective medicine is not a consumer protection violation, and a recent order in the Northern District of Illinois demonstrates that.  In Camargo v. AbbVie, Inc., No. 23-cv-02589, 2026 WL 115068 (N.D. Ill. Jan. 14, 2026), the district court dismissed a multistate class action alleging consumer protection claims for multiple reasons, but mainly because the plaintiffs got exactly what they paid for—even if they paid more than they would have liked for a life-improving product.  The court also ruled that federal patent law impliedly preempted the plaintiffs’ claims. 

In Camargo, residents of California, Connecticut, Indiana, and Michigan alleged that Humira’s list price was artificially inflated through a rebate-driven strategy with pharmacy benefit managers, thus forcing consumers who paid list price (or coinsurance based on list price) to bear “oppressive” costs.  They pleaded claims under the Illinois Consumer Fraud and Deceptive Business Practices Act (“ICFA”) and similar statutes in nearly thirty states.  The manufacturer moved to dismiss. 

To start, because the Illinois statute has no extraterritorial effect, the relevant transactions must have occurred “primarily and substantially” in Illinois.  But these plaintiffs lived out of state, and although they alleged that the pricing strategy was conceived and implemented in Illinois, that was not enough.  All the plaintiffs filled their prescriptions in other places, so the court dismissed the ICFA claims brought by these carpetbagging out of staters.  

Separately, the court found that two plaintiffs failed to plead damages under the ICFA.  Paying an allegedly “oppressive” price for a product, without more, is insufficient to establish a consumer injury, where the plaintiffs do not allege the product was worth less than the price paid or that they could have obtained a lower price elsewhere.  One plaintiff never paid the post-insurance “list” price, and another similarly failed to allege that Humira was defective or worth less than what he paid.  These failures provided independent bases to dismiss their ICFA claims.

There were more problems with the plaintiffs ICFA claims, namely the three-year statute of limitations.  Claims were time-barred to the extent premised on prices paid through 2018, when the plaintiffs purchased the product and their claims accrued.  Moreover, the “discovery rule” did not save the claim because the allegedly unfair prices were apparent at the time of payment, even if plaintiffs alleged did not know the defendant’s pricing strategy.

Beyond Illinois, recall that these plaintiffs were from California, Connecticut, Indiana, and Michigan.  But the complaint asserted no claims under Indiana or Michigan law, and the plaintiffs largely failed to defend other states’ claims, effectively waiving them. 

The court therefore focused on the California and Connecticut statutes, ultimately finding the allegations insufficient under each.  Under California’s Unfair Competition Law, plaintiffs alleged an “unfair” business practice, but the complaint did not plausibly allege anticompetitive effects.  To the contrary, it reflected the availability of alternatives and biosimilars.  Indeed, the California plaintiff switched to a biosimilar—undermining assertions of market foreclosure or harm to competition.  Under the Consumer Legal Remedies Act, a claim requires damages “as a result of an unlawful act.”  These plaintiffs alleged conduct that they did not like, such as alleged profit-maximizing and rebate practices they considered opaque.  But they alleged nothing unlawful.  The court dismissed both counts. 

The court found nothing wrong under the Connecticut Unfair Trade Practices Act either.  Under the CUTPA, “unfairness” is assessed by whether conduct falls within established concepts of unfairness, is immoral/oppressive, or causes substantial injury not outweighed by countervailing benefits and not reasonably avoidable by consumers.  Here, the Connecticut plaintiff in fact avoided paying Humira’s allegedly high prices, first through her insurance coverage and later by quitting the medication.  Even so, the plaintiffs alleged no statutory violation because “in any event, charging consumers higher prices, by itself, does not violate the CUTPA.”  Camargo, at *6. 

Most consequentially, the court held that plaintiffs’ pricing-based consumer claims were preempted by federal patent law to the extent they seek to penalize the manufacturer for charging “excessive” prices for a patented drug, citing Biotechnology Industry Organization v. D.C., 496 F.3d 1362 (Fed. Cir. 2007).  The court reasoned that penalizing high prices limits the full exercise of the exclusionary rights secured by patent law and therefore conflicts with congressional objectives, even though states generally may regulate unfair practices.

The plaintiffs cited the EpiPen MDL, where sales and antitrust claims were not preempted.  But their reliance on that case was unavailing because it involved allegations of monopolization conduct, deceptive marketing, and racketeering that independently supported state-law claims without targeting patent-derived pricing power.  Here, plaintiffs did not allege antitrust violations, exclusionary conduct against biosimilars, or deception.  To the contrary, they alleged that the manufacturer published its list prices, and at least two named plaintiffs used biosimilars, underscoring the absence of exclusionary practices.  As framed, the claims attack prices stemming from patent rights—squarely implicating patent law.

The court granted the manufacturer’s motion to dismiss on all claims, but without prejudice.  On this score, the court gave the plaintiffs until February 11, 2026, to file an amended complaint, but also cautioned that plaintiffs could do so if “consistent with their obligations under Rule 11.”  Camargo, at *9.  This court is clearly skeptical, for good reason. 

Photo of Bexis

We’ve only discussed Shady Grove Orthopedic Associates, P.A. v. Allstate Insurance  Co., 559 U.S. 393 (2010), a couple of times.  Shady Grove, displaced – in federal court – a variety of state-law limitations on class actions because those restrictions were at odds with Fed. R. Civ. P. 23, and in federal court federal rules properly enacted under the Rules Enabling Act, 28 U.S.C. §2072 (“REA”), control.  Thus, the plaintiffs in Shady Grove could bring state-law class actions in federal court that were barred in state court.

Earlier this week, in Berk v. Choy, 2026 WL 135974 (U.S. Jan. 20, 2026), a unanimous Supreme Court struck again, holding that a relatively common state-law restriction on medical malpractice claims – a requirement that the complaint be accompanied by an “affidavit of merit” signed by a doctor – did not apply in federal court because that requirement was beyond what was needed to satisfy Fed. R. Civ. P. 8.

The Court’s analysis in Berk seems equally relevant to the requirements of any federal rule.  First, the “Rules of Decision Act [28 U.S.C. §1652] dictates that state substantive law must yield if . . . a [federal] statute “otherwise require[s] or provide[s].”  Berk, 2026 WL 135974, at *3.  Second, the REA is such a statute, since it “authorizes“ the Supreme Court to “prescribe general rules of practice and procedure and rules of evidence,” §2072(a), “for district courts, [and] provides for the application of federal law.”  Berk, 2026 WL 135974, at *3 (citations omitted).  Thus, “a valid Rule of Civil Procedure displaces contrary state law even if the state law would qualify as substantive under Erie’s test.”  Id. (emphasis added).  As will become clear, we think that’s important.

Think about that again.  The Supreme Court just unanimously held that anything in state law, whether “procedural” or “substantive” goes out the door in federal diversity cases, to the extent that it “demands more,” 2026 WL 135974, at *4, than a federal rule – unless the rule itself exceeds the scope of the REA.  Id.

The state affidavit-of-merit requirement demanded more than Rule 8, since that rule has no requirement to include “evidence” (the affidavit) in a complaint.  Id. (“Under Rule 8, factual allegations are sufficient, but under the Delaware law, the plaintiff needs evidence too.”).  “The two rules thus give different answers.”  Id.  Nitpicking – that the affidavit was “a separate sheet of paper” from the complaint itself – was unsuccessful.  Id.  What matters is whether the state requirement and the applicable federal rule “address[] the same issue – and in doing so, impose[] a different standard.”  Id. (footnote omitted).  A number of more case-specific arguments made by the malpractice defendants also failed.  Id. at *5-6.

Those defendants’ final fallback position fared no better.  No, Rule 8 did not violate the REA because it was improperly “substantive.”  Id. at *6.

For purposes of the Rules Enabling Act, we use a modest test:  whether the Federal Rule really regulates procedure.  Or put differently, what matters is what the Rule itself regulates. . . .  In applying this analysis, we have rejected every statutory challenge to a Federal Rule that has come before us.

Berk, 2026 WL 135974, at *6 (citations and quotation marks omitted).  Berk had no trouble determining that Rule 8 – enacted as a Federal Rule of Civil Procedure – “really regulates procedure.”  Id.  While it had “practical effect on the parties’ rights,” it did not regulate “the rights themselves.”  Id.

Nor was Berk willing to change the Court’s long-standing refusal to evaluate the nature of the state law being displaced:

[Defendants] argue that determining whether a Rule is valid under the Rules Enabling Act requires asking a second question:  whether the displaced state law is substantive.  We rejected that approach eight decades ago and decline to reconsider it now.  On the contrary, we underscore that “the substantive nature of [a state] law, or its substantive purpose, makes no difference.

Berk, 2026 WL 135974, at *7 (citations omitted) (emphasis original).

So what can we take away from Berk?  Start with medical malpractice, the subject of the state legislation overturned (preempted?) in Berk.  Most p-side lawyers specializing in such cases wouldn’t be caught dead in federal court if they could help it.  Does that change?  Maybe some in the short term, but probably not in the long term.  Most malpractice cases are non-diverse to start with, and the sorts of malpractice plaintiffs likely to have trouble satisfying certificate-of-merit requirements don’t have very good cases anyway, so we question how many will want to “make a federal case” of it.

On the other side, if the states have the political will to do so, it wouldn’t be hard to fix the federal rules problem that did in this particular certificate-of-merit statute.  Just tie the requirement to something that the federal rules don’t reach.  A certificate/affidavit mandate as part of initial discovery might run afoul of Fed. R. Civ. P. 26(a)(2) governing initial disclosures.  So tie such mandates to something else.  The statute of limitations for medical malpractice could be 180 days, unless the plaintiff submits a certificate of merit, in which case it’s the same length it was before.  Or, because medical malpractice cases without certificates of merit are highly likely to be bogus, the damages cap for such cases is very low, whereas with the certificate, it is whatever state law now requires.  Because an amended certificate/affidavit mandate no longer has anything to do with anything a federal rule covers, it should have no trouble with Berk.

What else?  Well, here’s an interesting piece discussing how Berk might cause trouble for parties in federal court seeking dismissal of litigation under state “Anti-SLAPP” statutes. Nor should state restrictions on informal interviews with plaintiffs’ treating physicians apply in federal court in derogation of broader federal discovery rules.

But we hasten to point out that Berk implicates more than just the Federal Rules of Civil Procedure.  As we quoted at the beginning of this post, the Federal Rules of Evidence also stem from the REA and apply in federal court under the Rules of Decision Act.  That’s why Rule 702 is so important in federal prescription medical product liability litigation.  But other state-law evidentiary peculiarities besides expert testimony could well conflict with the Federal Rules of Evidence – particularly the liberal admissibility requirements of Fed. R. Evid. 401 and 402.  Rule 401 provides that “Evidence is relevant if . . . it has any tendency to make a fact more or less probable than it would be without the evidence.”  Rule 402 provides that “[r]elevant evidence is admissible” subject to a number of exceptions that do not include any form of state law.

In a prior post from a couple of years ago, we discussed the original Shady Grove decision in the context of a couple of ways that Pennsylvania evidentiary restrictions in product liability cases could be trumped by the broader admissibility standards of Fed. R. Evid. 401-02.  Berk only reinforces those arguments, with its repeated emphasis that, when federal rules are involved, whether the state law being displaced is “substantive” or “procedural” doesn’t matter a hill of beans.  Subsequent Pennsylvania developments only increase the importance of applying the Federal Rules of Evidence in federal court.  Since that post, the Pennsylvania Supreme Court has made Pennsylvania the only state in the nation where compliance with industry and governmental (in our sandbox read, FDA) standards is not even admissible in strict liability actions.  See Sullivan v. Werner Co., 306 A.3d 846, 862-63 (Pa. 2023) (plurality opinion), affirming, 253 A.3d 730, 747 (Pa. Super. 2021) (discussed in detail here).

Under Berk and Fed. R. Evid. 401-02, that evidentiary restriction should not apply in federal court.  And it hasn’t, even before Berk.  The Third Circuit has repeatedly held that the Federal Rules of Evidence govern in diversity product liability cases otherwise subject to state law.  E.g., Covell v. Bell Sports, Inc., 651 F.3d 35, 36-37 (3d Cir. 2011); Moyer v. United Dominion Industries, Inc., 473 F.3d 532, 546 (3d Cir. 2007); Diehl v. Blaw-Knox, 360 F.3d 426, 431 & n.3 (3d Cir. 2004); Rolick v. Collins Pine Co., 975 F.2d 1009, 1013 (3d Cir. 1992):

The issue to be decided here is whether the OSHA regulation is admissible in a diversity action as evidence of the standard of care owed by the defendants to the plaintiff. . . .  Since the question involves the admission of evidence in a federal court, the Federal Rules of Evidence control. . . .  We can think of no reason under the Federal Rules of Evidence why the OSHA regulation is not relevant evidence of the standard of care once it is determined, as we have done, that under Pennsylvania law the defendants could owe plaintiff a duty of care.

Rolick, 975 F.2d at 354.  See Kelly v. Crown Equipment Co., 970 F.2d 1273, 1278 (3d Cir. 1992) (strict product liability case; federal rule admitting subsequent remedial measures “is ‘arguably procedural,’ and therefore governs in this diversity action notwithstanding Pennsylvania law to the contrary”). NOTE: none of these cases involve prescription medical products. That’s because Pennsylvania has strictly enforced Restatement (Second) of Torts §402A, comment k (1965), and not applied strict liability to prescription products.

Berk only reinforces this Third Circuit precedent, so in Pennsylvania federal court product liability litigation, compliance evidence should be admissible regardless of Sullivan.  Thus, Third Circuit Rule 401-02 decisions in product liability cases applying Pennsylvania substantive law should apply, notwithstanding Sullivan. Federal standards on what‘s “relevant,“ rather than Pennsylvania’s exclusion of compliance evidence based on its peculiar (and probably “substantive”) reading of Restatement §402A.  As to the admissibility of compliance, Pennsylvania law plainly “demands more” than Rules 401-02 and when applied to compliance evidence would “give different answers” to admissibility.  Berk, 2026 WL 135974, at *4.  Since it can hardly be denied that Rules 401-02 “really regulate” evidence, that Pennsylvania might consider its more restrictive approach “substantive” is of no consequence.  Id. at *6.

To the extent that states have other restrictive evidentiary rules that disfavor defendants (statutes barring admission of seatbelt nonuse come to mind), a similar rationale should apply after Berk.

Photo of Eric Alexander

Not quite three years ago, we co-authored a chapter in a Digital Health guide put out by International Comparative Legal Guides.  It bore the pithy title “Predicting Risk and Examining the Intersection of Traditional Principles of Product Liability Laws with Digital Health.”  We continue to tinker with the principles of product liability law and monitor how they intersect with digital health, software, artificial intelligence, and other aspects of our increasingly abstract existence.  Our then-colleague Gerry Stegmaier with whom we co-authored the chapter still lives in the world of companies that develop software and AI platforms, so we thought it made sense to get his views again.  Perhaps because January invites retrospection (and because high tech years have at least a canine multiplier), we decided to see how we did with our predictions.  Being lawyers, our prior predictions were not overtly identified as such, but it is still possible to pull out the forward-looking statements we made with a probabilistic bent.

Our declared idealistic “aim to aid software developers in digital health and those that advise them in anticipating, preparing for, and responding to this potentially rapid changing liability landscape” was offered in the context of four general developments over the few years prior to early 2023.  First, many digital health products, medical devices and otherwise, were being developed, often by companies recently created to tackle digital health needs.  Second, a number of decisions had come out that tested the presumptive treatment of software as a licensed service or intangible, not a product.  See here, here, and here.  Relatedly, the creation of the high profile Social Media MDL (In re: Social Media Adolescent Addiction/Personal Injury Products Liability Litigation) meant that these issues were going to get increased attention.  Third, FDA had been in catch-up mode, issuing a number of guidance documents about software and medical applications that were arguably long overdue.  (The history of FDA regulation in this area is included in the second section here.)  Fourth, although not the focus of our chapter, the writing was on the wall that the EU was going to issue a directive treating software as a product for purposes of liability, which it ultimately did in October 2024.  While the U.S. does not tend to follow what the E.U. does on product liability, companies that hope to sell their products or license their technologies around the world do have to pay attention to such things. 

Taking our predictions in the order in which they appeared in the chapter, the first contains an implicit prediction:  “As healthcare continues to become more digital, the prevalence of devices reliant on software continues to grow.”  This was low hanging fruit, so to speak, but both trends have continued.  Public lists of FDA approvals and clearances since early 2023 include many software-driven devices.  In the “Device Innovation” section of its 2024 Annual Report, FDA’s Center for Devices and Radiologic Health identified eight examples of novel devices, of which four were software and two others utilized software.  We find that telling.

Next, we identified “a number of areas where the application of traditional U.S. product liability principles could differ with patient-facing digital health and software-driven medical devices.”  (Our apologies in advance for any confusion over the British spellings that carry over from the chapter in International Comparative Legal Guides.)

  • We noted the possibility for liability based on a post-sale duty to warn for digital health and software-driven devices “where the relationship with the end-used may continue post-sale and the ability to update software may go hand-in-hand with the ability to notify an end-user of a post-sale issue.”
  • We noted that software update abilities with prescription medical devices could mean that “the learned intermediary doctrine may not apply in some cases where the level of direct and/or continuing interaction between the manufacturer and patient undercuts the rationale for the doctrine.”
  • We postulated a poor preemption result for a PMA device when, “[i]n the case of a device with software that will be updated over time or where the device utilises AI or machine learning, some courts may doubt that the device at the time of the alleged injury was the same as what FDA had approved.”
  • “Lawsuits over injuries allegedly due to a failure of software in a medical device might also name entities that contracted with the device manufacturer to develop or update that software. A parallel may be seen in the history of suing manufacturer.”
  • We suggested that federal legislation could try to impose “limitations on liability” if civil litigation impeded innovation.
  • We also envisioned courts creating easier routes to liability without proof of defect or negligence in the situation where “software fails to perform as expected and there is no ability for it to be altered by anyone other than the manufacturer or its agents.”
  • Suits alleging product liability in the absence of “tangible physical injuries,” including under increased risk or medical monitoring theories.
  • “Given that product liability law has generally not applied to software, any imposition of product liability would entail making new state law.”

Addressing each of these in turn would make this post far too detailed and long to keep any reader’s attention, so we will summarize them and focus on a few.  The summary is that positions taken in litigation, decisions issued by courts, and who is getting sued over digital health, including devices driven by AI and software, are in flux.  We have been tracking them here and in other specific posts.  Litigation in this space has increased, but the relatively few decisions since early 2023 have not yet established any consensus.

On our point above about “new state law,” our chapter specifically looked to the Social Media MDL, finding it “highly likely that the issue of whether strict product liability applies to social media platforms, including the software that runs them, will be decided directly. Those decisions, potentially modified on appellate review, will inevitably influence the legal playing field for potential claims relating to digital health and software-driven medical devices.”  We do have decisions from that court, including In re Social Media Adolescent Addiction/Personal Injury Products Liability Litigation, 702 F. Supp. 3d 809 (N.D. Cal. 2023), discussed here, but we do not yet have a clear articulation of an adoption of rejection of new law for specific states.  Without speaking to the law of specific states consistent with Erie, that court denied a motion to dismiss asserted product liability claims because “plaintiffs adequately plead[ed] the existence of product components as to each alleged defect analyzed” under a novel approach focused on “functionality” instead of “tangibility” or other traditional measures of what constitutes a “product” for purposes of product liability.  We cannot say that this novel way of defining what is a product will have any legs,  Later, in the same MDL in the context of plaintiffs pushing public nuisance under what is fundamentally a product liability theory, the plaintiffs denied that they were asserting product liability claims at all and the court characterized the issue of whether the social media platforms were products as disputed.  Clearly, this is going to require more litigation, especially on appeal, to get some clarity.

As for our prediction on federal legislation, the pending AI Lead Act is certainly broader than a software version of the Biomaterials Access Assurance Act.  If enacted, it would create a federal cause of action, preempt some state laws, and definitely treat software as a product.  Until it is passed in some form, though, it is unclear how this will change the playing field.  The history of product liability litigation over PMA devices and other medical products with applicable express preemption provisions teaches that plaintiff lawyers will make a concerted and prolonged effort to erase any statutory limits on liability and damages.  We cannot help but see the chilling effect that the plaintiff bar’s on-going efforts to expand liability, both in terms of novel claims and new targets, can have on investment in developing new technologies in digital health.

Speaking of prolonged efforts, we also predicted that FDA would be issuing more guidance on software as devices and devices driven by software, even though it viewed its authority in this area as limited.  In its September 2022 Policy for Device Software Functions and Mobile Medical Applications, FDA promised to “continue to evaluate the potential impact these technologies might have on improving health care, reducing potential medical mistakes, and protecting patients”  Earlier this month, made good on its promise with two new guidances, which replaced guidances from September 2022 and September 2019, respectively, and a new draft guidance.  Briefly—because each could have a deep dive on its content and potential liability implications—the guidance on Clinical Decision Support Software reexamined “FDA’s oversight of clinical decision support software intended for health care professionals . . . as devices,” which is undoubtedly one of the hot development areas.  A big part of the guidance on General Wellness: Policy for Low Risk Devices is to help define which software is a device and which is merely an application to help people stay or become healthier without treating a specific disease.  FDA does not regulate the latter, so certainty on what is and is not a device is important but perhaps ephemeral.  The draft guidance Artificial Intelligence-Enabled Device Software Functions: Lifecycle Management and Marketing Submission Recommendations builds on a number of FDA statements over the last 19 months.  See, e.g., here, here, and here.  Draft guidances are subject to public comment, and FDA responses to those can be informative.  Sometimes, FDA’s draft guidances have stayed that way for years, taking on the same functional authority as final guidances, which are still considered “nonbinding.”  So, the short- and long-term impacts of this draft guidance remain to be seen. More generally, given that these three (draft) guidances were issued within four weeks of an Executive Order about setting a national policy on AI, it seems inevitable that the regulation of software as devices and devices driven by software will continue to be an area of change for the foreseeable future.

Our last two predictions were, frankly, pretty obvious in hindsight.  We said:

[S]oftware-development lifecycle best practices are likely to evolve further and familiarity with FDA’s risk classification schemes may be a useful starting point for many developers, regardless of whether their software is or may be a medical device.

***

In any event, medical device companies, software developers who work with them, and those who assist each with managing and responding to liability risks will benefit from greater understanding of and monitoring this emerging area of the law.

We continue to think these are sound predictions, even if made in early 2026.  In this space, risk minimization cannot be driven solely by traditional views about software—i.e., contractual limits were—or doctrinaire views about product liability.  We expect that digital health will continue to be area where legislators, regulators, litigators, and judges all play a role in setting the ground rules that will emerge over time.  As that happens, companies in this space have to take steps to minimize risks even while ambiguity and uncertainty remain about what those ground rules will be.  It would be a shame if the uncertainty hampers the development, deployment, and adoption of AI-powered products and services that will advance public health and the delivery of healthcare. 

Drug & Device Resources

Bexis’ Book

FDA-regulated products now account for an estimated one-fifth of overall economic activity in the U.S. They have also been the focus of a litigation explosion. This timely guide covers all aspects of litigation involving drugs, medical devices, vaccines and other FDA-regulated prescription products.
Learn more about this book