OpenAI Faces Lawsuits From Multiple States Over Claims ChatGPT Harms Children and Deceives Adults.
U.S. states are suing OpenAI, accusing the company behind ChatGPT of putting profits ahead of public safety, exposing children to dangerous conversations, and misleading adults about the risks of artificial intelligence.
The lawsuits allege that ChatGPT has encouraged suicidal behavior, provided potentially dangerous medical advice and reinforced paranoid thinking among vulnerable users.
Prosecutors also accuse OpenAI of failing to adequately control advanced AI systems capable of deceptive behavior and unauthorized cyberattacks. The filings identify Arizona, Arkansas, Idaho, Kansas, Montana, Nebraska, and West Virginia as the states in the lawsuit. Iowa Attorney General Brenna Bird also publicly condemned the company, bringing the number of states involved in the reported legal action to at least eight.
“OpenAI is hurting kids and they are lying to Iowans,” Bird said in a statement. According to the lawsuits, ChatGPT can create the impression of an emotionally supportive relationship while producing responses that may reinforce harmful thoughts or discourage users from seeking help from people around them.
Arizona’s complaint specifically alleges that the chatbot encourages teenagers to disclose personal information and sometimes discourages them from sharing sensitive conversations with parents, friends, or authorities.
The states argue that OpenAI knew, or should have known, about these risks but continued expanding its products while presenting its safety testing and protections as adequate.
Idaho’s lawsuit goes further, accusing the company of exposing the public to consequences ranging from personal tragedies to cybersecurity threats.
“Those revenues have come at great cost to the public,” the complaint states, citing deaths, cyberattacks and broader risks associated with advanced AI systems. The complaints also describe allegedly rogue AI agents capable of hacking other companies, deceiving human operators and changing their own instructions.
OpenAI has disputed the suggestion that such behavior represents an uncontrolled threat, saying it investigates misaligned models and has not identified anything comparable to a previously reported July incident involving an AI agent.
However, the effectiveness of those safeguards is already being challenged. Common Sense Media, a nonprofit organization that evaluates technology’s impact on children, reported last week that some protections in ChatGPT for Teens failed during its testing. The organization recommended restricting access to adults until stronger safeguards can be demonstrated.
The company also says independent experts review its policies and chatbot behavior.
The lawsuits follow an $18 billion settlement involving Meta and claims related to teenage social media addiction. That agreement, reached in August, included additional child-safety measures for Facebook and Instagram.
Matthew Bergman, founding attorney of the Social Media Victims Law Center, described the OpenAI cases as “the most significant coordinated state action ever taken against an artificial intelligence company.”
