US State Attorneys General Pursue Product Liability Lawsuits Against AI Chatbot Companies, Echoing Big Tobacco Litigation Strategy
A coordinated litigation strategy modelled on the legal campaigns that defeated the tobacco industry is being deployed against artificial intelligence companies in the United States, with a state attorney general filing suit alleging that AI chatbots cause mental health harms to users. The move, described as the first attorney general action of its kind against AI chatbot companies, parallels the product liability approach already generating thousands of cases against major social media platforms. The suits rely on product liability law — the legal framework that holds manufacturers responsible for foreseeable harms caused by their products — applied to AI-generated outputs and chatbot interactions. Critics of AI regulation at the federal level, including Michelle Lopes Maldonado of the Information Technology and Innovation Foundation (a think tank whose funders include AI developers such as Anthropic and Alphabet), have acknowledged that this litigation reflects "frustrations that there's not a uniform, cohesive standard at the federal level." The absence of a comprehensive US federal AI liability framework is doing what it did for social media before federal action: driving a fragmented, state-level litigation environment where plaintiffs use existing tort and product liability doctrines to create de facto regulation through the courts. For AI companies operating in or selling into the UK and EU, where the EU AI Act (the European Union's risk-based regulatory framework for artificial intelligence, adopted in 2024) is being implemented, the US litigation trajectory adds a further layer of liability exposure to consider alongside statutory compliance obligations.
Why this matters
The application of product liability doctrine to AI chatbots is a genuinely novel legal development with direct implications for AI governance and technology law practices. If state-level product liability suits gain traction — as they did against social media platforms — AI developers will face mounting pressure to redesign systems, adopt safety labels, and create litigation-defensible governance frameworks. For UK and EU practitioners, the US litigation arc serves as a leading indicator: the EU AI Act already imposes risk-classification and transparency obligations on certain AI systems, and UK courts may be asked to consider analogous product liability claims under domestic law as AI adoption grows. This story directly implicates technology transactions, product liability, and AI regulatory advisory work.
On the Ground
A trainee advising an AI company client on this developing litigation landscape would be conducting AI governance policy drafting to document safety measures and content moderation frameworks, completing vendor due diligence questionnaires assessing third-party AI tool risks, and preparing regulatory impact assessment memos comparing the EU AI Act's requirements against the company's current deployment practices. Reviewing data processing agreements and technology licence terms for liability allocation clauses would also be active tasks.
Interview prep
Soundbite
Product liability suits against AI chatbots are creating US case law that will land on EU and UK desks within two years.
Question you might get
“How does product liability law apply to harms caused by AI chatbot outputs, and what defences might an AI developer raise in response to such a claim under English law?”
Full answer
A US state attorney general has filed what is described as the first government-level product liability lawsuit against an AI chatbot company, alleging mental health harms to users — a legal strategy consciously modelled on the Big Tobacco litigation that transformed the tobacco industry's liability landscape. The suit matters because product liability doctrine is jurisdiction-portable: if US courts accept it for AI outputs, English and European courts will be asked the same questions under domestic equivalents. The EU AI Act already imposes obligations on high-risk AI systems, and UK product safety law could be read to cover AI-generated outputs causing foreseeable harm. My view is that this will accelerate demand for AI governance advisory and product liability pre-litigation counselling as AI companies seek to build litigation-defensible safety architectures before the wave reaches European courts.
My notes
saved