USA Trends and Developments Contributed by: Gregory Ulmer, James Phillips and Ryan Walton, BakerHostetler
In In re Social Media Adolescent Addiction / Personal Injury Prods . Liab . Litig ., hundreds of actions brought by minors, parents, school districts and state attor - neys general were consolidated in the Northern Dis - trict of California. Plaintiffs alleged that major social media platforms (including Meta, Google/YouTube, TikTok and Snap) were designed to foster compul - sive use by adolescents, causing mental health and physical harms. Id at 865–66. Critically, plaintiffs did not base their core product liability theories on third-party content. Instead, they challenged specific platform features and user-expe - rience design choices, including: • endless-scroll feeds; • algorithmic content prioritisation; • push notifications and reward mechanics; • lack of age verification and parental controls; and • default settings that allegedly promoted compul - sive use. Defendants moved to dismiss, arguing (among other things) that: • social media platforms are services , not products; and • product liability law therefore could not apply as a matter of law (Id). Where the social media addiction multi-district liti - gation addressed large digital platforms, Garcia v Character Technologies applies the same reasoning to a generative-AI chatbot, confirming that courts are extending function-focused, design-centric analysis beyond social media and into standalone AI sys - tems. See Garcia v Character Technologies , Inc , 785 F.Supp.3d 1157, 1180 (M.D.Fla., 2025). The court denied the defendants’ motion to dismiss the strict product liability design-defect claim, explicit - ly rejecting the defendants’ categorical argument that software and AI tools are outside product liability law as a matter of definition. Id. Critically:
• the court distinguished between content and design, holding that the claims targeted the archi- tecture and operation of the chatbot, not the expressiveness of individual messages; • the court emphasised that the alleged harm arose from how the system was designed to function and interact with users, including defaults, reinforce - ment loops and lack of safeguards; and • the court treated the chatbot as a consumer prod - uct for purposes of strict liability, to the extent that the claims arose from alleged design defects rather than ideas or expression (Id). Both cases reject categorical defences and instead ask whether the design behaves like something that product liability law already regulates. Rather than ask whether software is a product in the abstract, courts ask whether particular features operate like traditional product components, safety mechanisms or control systems. Under this feature-specific, design-centred approach, plaintiffs can advance product liability theories without winning a global definitional fight. The enquiry shifts to whether the alleged defect mirrors design choices that product liability law already evaluates: warnings, defaults, safeguards, foreseeable misuse, and alterna - tive designs. This shift creates a practical asymmetry. Defendants may ultimately succeed on doctrinal grounds and still incur significant exposure well before trial, including expansive discovery into design decisions, inter - nal testing and risk assessments. Even successful motions may arrive only after reputational damage, adverse precedent or settlement pressure driven not by liability probability but by verdict volatility. This functional turn reflects a broader judicial discom - fort with bright-line formalism in the face of rapidly evolving technology. Courts are increasingly aware that traditional product/service distinctions were developed for a physical-goods economy and can obscure, rather than illuminate, how modern tech - nologies actually operate and create risk. As digital systems assume roles once filled by tangible products – guiding decisions, shaping behaviour and mediating safety-critical interactions – courts have become less
329 CHAMBERS.COM
Powered by FlippingBook