Insights & Events

Two verdicts handed down on back-to-back days in March 2026 may prove to be as consequential for the technology industry as any regulatory action in recent memory. On March 24, a New Mexico jury found Meta liable under the state’s Unfair Practices Act for deceiving consumers about the safety of its platforms for children, awarding $375 million in civil penalties. The following day, a Los Angeles jury found Meta and Google had each knowingly designed dangerous products, Instagram and YouTube, and allowed minors access without adequate warning, awarding $3 million in compensatory damages and $3 million in punitive damages to an individual minor plaintiff – identified by the initials K.G.M.

For companies that build, operate, or invest in consumer-facing digital platforms, these verdicts, if they stand, demand attention as signals of a potential shift in how some courts approach the application of products liability principles to digital media platforms.

The Section 230 End-Run: Design Defect as the New Theory of Liability

For years, social media companies relied on 47 USCS § 230 of the Communications Decency Act (“Section 230”) to defend products liability claims brought against them. Section 230 broadly immunizes online platforms from liability for content posted by third-party users. Historically, courts have often treated software as a service and have been reluctant to apply products liability theories to claims that software caused an injury. Plaintiffs pursuing products liability claims against online platforms, however, have argued that Section 230 does not bar claims that the platforms themselves are defective. Plaintiffs have argued, and some courts have agreed, that Section 230 protects platforms from claims rooted in what users say, not in how the platform itself was built. For example, the Massachusetts Supreme Court recently ruled on April 10, 2026, that Section 230 did not bar the state’s claims that Meta’s platform was addictive to teens. See Commonwealth v. Meta Platforms, Inc., 497 Mass. 384, 277 N.E.3d 166 (Mass. 2026).

Following this theory, plaintiffs in K.G.M. v. Meta and YouTube argued that there is a distinction between the platform itself and the content. Rather than challenging any specific item of content, plaintiffs focused on elements of the architecture of the platforms themselves – features like infinite scroll, algorithmic content amplification, autoplay, and push notifications. The plaintiffs argued that these features were deliberately engineered to maximize engagement at the expense of user safety, particularly for minors. With the court disallowing a defense based on Section 230, the jury found that the companies knew their design choices were likely dangerous and failed to warn users. Following the verdict, Meta reasserted the Section 230 defense in their argument for judgment notwithstanding the verdict. Meta argued that the plaintiff’s addiction stemmed from her interaction with third-party content on their platform, such as messages, likes, and photos, rather than the platform’s features and that these features merely facilitate the publication of third-party content, which is protected under Section 230. Meta separately argued that the First Amendment independently shields its editorial decisions about how content is curated and displayed on its platforms, relying on Moody v. NetChoice, LLC, 603 U.S. 707 (2024), which found “the First Amendment offers protection when an entity engaging in expressive activity, including compiling and curating others’ speech, is directed to accommodate messages it would prefer to exclude.” Id. at 731. And “[t]o the extent that social-media platforms create expressive products, they receive the First Amendment’s protection.” Id. at 716. The Los Angeles court denied Meta’s motion.

New Mexico’s attorney general in State of New Mexico v. Meta Platforms, Inc. alleged that Meta designed and operated its platforms with inadequate protections that enabled predators to find, groom, and exploit vulnerable children. The jury found Meta liable for violating New Mexico’s Unfair Practices Act by engaging in “an unconscionable trade practice” and awarded $375 million rather than the $2 billion suggested by the state. A separate bench trial on a public nuisance claim concluded, but the court has not issued a decision. Meta has announced its intent to appeal the jury verdict.

Upcoming Trials

These are not isolated cases: there are more than 800 school district claims and 10,000 individual personal injury cases (over 2,500 of which are pending before a single judge in a federal multidistrict litigation in the Northern District of California), and cases brought by state attorneys general. All allege that social media platforms were engineered to addict children and teenagers. See MDL Statistics Report – Distribution of Pending MDL; MDL Update: Social Medial Adolescent Addiction Lawsuit.

We will be monitoring how litigation in this space proceeds. In addition to the anticipated appeals mentioned above, the first MDL bellwether trial was set to begin on June 15, 2026, but resolved days before trial. The next bellwether is scheduled to begin August 6, 2026, and involves claims brought by various state attorneys general. This trial, and the judicial decisions that come out of it, will shape the trajectory of the litigation and provide further insights into how juries evaluate causation and damages in these cases.

What Technology Companies Should Be Doing

The March 2026 verdicts do not just affect social media companies. They reflect a broader judicial willingness to apply traditional products liability principles such as design defect, failure to warn, and strict liability to digital products and AI systems. Recent litigation already tests whether AI chatbots can be treated as consumer products and what, if any, duties developers owe when those products are accessed by vulnerable users.

For example, in Garcia v. Character Techs. Inc., brought in federal court in the Middle District of Florida, plaintiffs asserted a wrongful death claim and alleged that an AI chatbot contributed to a minor’s suicide by generating harmful and manipulative responses. The court’s application of the Restatement (Third) of Torts’ “reasonable alternative design” test to adaptive, emergent software systems raised novel questions about how courts will evaluate the safety of AI-driven platforms and how to apply First Amendment protections to “speech” generated by AI. Answers may be found in future cases, as the parties announced they reached an agreement in principle to resolve the claims in Garcia.

Other key questions remain, such as: how will courts evaluate Section 230 defenses and the “product vs. service” distinction as applied to other types of digital platforms and AI-driven algorithmic systems? Will the design defect framework extend to platforms used primarily by adults? And how will courts handle the issue of causation, the demanding task of tying a specific platform design feature to a specific plaintiff’s mental health outcome?

For technology companies, several near-term priorities emerge. First, companies facing potential litigation should keep in mind that certain corporate documents will be discoverable in litigation. Undertaking an internal analysis to ensure the documents tell the full and accurate corporate story is essential. Further, the evolving MDL landscape requires early coordination strategy. With thousands of cases pending, bellwether trial selection and discovery management will be extremely important. The plaintiffs’ bar advances their favored cases to the front of the line for early trials – and those cases are rarely representative of the broader pool. Aggressive and coordinated defense efforts to ensure that representative cases are being tried (and meritless cases dismissed) will be critical.