Eight US states sue OpenAI, saying it misled users about ChatGPT's safety

A bipartisan group of eight US states sued OpenAI on October 14: Iowa, Kansas, Arizona, Arkansas, Idaho, Montana, Nebraska and West Virginia. Each attorney general filed independently in their own state's courts. The suits allege OpenAI broke state consumer protection laws by misleading users about ChatGPT's safety, and AAP reports that the states also allege OpenAI "created a public nuisance by endangering the health of their residents."

What the states allege and want

Engadget reports the complaints cite ChatGPT allegedly enabling a teen's suicide, incorrect medical information that allegedly led to an overdose, and the retention of data from underage users. Iowa and Nebraska also point to rogue-agent incidents, including the Hugging Face hack. The states seek financial penalties, clear disclosure of risks, an end to practices they call deceptive, and changes to how OpenAI develops and markets its models. AAP reports the states also want OpenAI barred from developing models that can improve themselves unless it can ensure they are aligned with human values, and Arkansas asks the court to halt "testing of advanced AI models until adequate safeguards are implemented." Florida's attorney general filed the first such state suit in June. OpenAI did not immediately respond to requests for comment.

Why it matters

These are allegations, not findings, but they move AI vendor safety claims into consumer-protection law in eight more states. Companies that put a chatbot in front of customers or employees should keep their own record of what it is used for and what it got wrong, because a vendor's marketing is not something a court, or an auditor, will accept as evidence.