A bipartisan coalition says “delusional” artificial intelligence responses could put users at risk and may violate state consumer protection laws.
A bipartisan coalition of 42 state attorneys general formally warned some of the nation’s largest artificial intelligence (AI) companies that their chatbot systems may be producing dangerous “delusional” outputs that could violate state law and harm users. The letter, issued Dec. 9 by the National Association of Attorneys General and signed by 42 attorneys general, was sent to 13 major AI developers, including Microsoft, OpenAI, Google, Meta, Apple, and Anthropic.
In the document, the officials say generative artificial intelligence (GenAI) “has the potential to change how the world works in a positive way,” but added that it also “has caused, and has the potential to cause, serious harm, especially to vulnerable populations.”
The letter underscores incidents in which AI interactions allegedly contributed to real-world harm, including suicides and violent acts, although it does not connect specific cases to individual companies. It expresses concern that some chatbots have provided responses that “either encouraged users’ delusions or assured users that they were not delusional.”
Signatories press for a range of safeguards. They want companies to post “clear and conspicuous” warnings about the risks of harmful AI outputs, adopt transparent third-party audits, and notify users when they are exposed to potentially dangerous responses. The attorneys general also urged firms to disconnect revenue incentives from safety decisions and treat mental health incidents with the same reporting rigor as cybersecurity breaches.
“We therefore insist you mitigate the harm caused by sycophantic and delusional outputs from your GenAI, and adopt additional safeguards to protect children,” the letter states, adding that failing to do so “may violate our respective laws.”
Iowa Attorney General Brenna Bird, one of the signatories, said in a statement that states are “putting the AI giants on notice: our children’s well-being will not be exploited for profit.”
The attorneys general set a Jan. 16, 2026, deadline for the companies to respond with commitments. Several firms have not publicly commented on the letter.