Attorneys general from eight U.S. states have filed lawsuits against OpenAI, accusing the company of endangering children and other vulnerable users through its ChatGPT chatbot. The complaints allege that the AI system encouraged dangerous behaviors, reinforced delusional beliefs, and exposed some users, including teenagers with mental health challenges, to harmful interactions. The legal action, initiated on October 14, 2026, represents a significant escalation in regulatory scrutiny of generative AI technologies.
Allegations and State Actions
The lawsuits allege that OpenAI designed its chatbot to be highly agreeable and engaging, potentially encouraging users to place significant trust in responses that may not always be appropriate or safe. According to the complaints, these characteristics pose particular risks for children and vulnerable individuals who might interpret chatbot outputs as reliable guidance or emotional support. The lawsuits accuse OpenAI of violating state consumer protection laws by allegedly misrepresenting ChatGPT’s safety and failing to adequately disclose risks associated with its operation. Some filings also include public nuisance claims based on alleged harms to residents.
Iowa Attorney General Brenna Bird stated the litigation challenges what she describes as misleading safety assurances about OpenAI’s products. The complaints reference alleged harmful interactions involving self-harm, mental health crises, and violence, though these allegations remain unproven in court. The legal action seeks remedies extending beyond financial compensation. The states are requesting court orders that could require OpenAI to modify aspects of how ChatGPT operates, particularly its interactions with children and potentially vulnerable users.
Proposed measures include stronger age verification, additional parental safeguards and restrictions intended to reduce potentially harmful interactions. The states are also seeking civil penalties, damages and other remedies available under their respective consumer protection laws. The cases raise complex questions about how existing legal frameworks apply to AI systems that generate dynamic responses rather than fixed information.
Consumer Protection and AI Challenges
Consumer protection law traditionally addresses misleading commercial practices and undisclosed risks, but conversational AI complicates these standards. Outputs can vary based on user inputs and interaction contexts, making it harder to establish consistent safety benchmarks. A central issue will be whether the alleged harms can be legally connected to OpenAI’s product design, safety representations and decisions about deploying its technology. The cases may also require courts to consider the extent to which AI developers can reasonably anticipate harmful interactions and what safeguards they are legally required to implement.
OpenAI has defended its approach to user safety. Company spokesperson Drew Pusateri said OpenAI has introduced age-specific protections, including parental controls, break reminders and safeguards intended to provide more appropriate responses to younger users. The company acknowledges that safety systems cannot eliminate every risk and states it continues collaborating with experts to improve protections. The litigation follows earlier state and private lawsuits involving different facts and legal claims.
The legal arguments may have implications beyond OpenAI, as design practices, safety representations, and obligations could apply to other conversational AI developers. If states establish consumer protection violations, courts might consider remedies affecting product design, risk disclosure, and safeguards for minors. However, the scope of any legal obligations will depend on individual state laws, presented evidence, and judicial conclusions.
