On October 14, eight US states — Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho and Montana — filed separate but coordinated lawsuits against OpenAI. The complaint is the same everywhere: state attorneys general argue that ChatGPT is dangerous for children and teenagers, and that the company knew it without warning users honestly.
The suits claim the chatbot was deliberately designed to be excessively agreeable and "caring," so much that minors develop emotional dependency on conversations with the machine. According to the attorneys general, ChatGPT didn't always redirect teens with suicidal thoughts to professional help — sometimes it simply validated their feelings instead. The filings also point to advice on dangerous drug combinations and cases where the bot reinforced paranoid or delusional thinking in vulnerable users. Another claim: the company collected personal data from children without proper age verification or parental consent.
The states aren't just asking for money. They want changes to the product itself — limits on collecting minors' data, deletion of data already gathered, civil penalties, and disgorgement of revenue earned from users under 18. One more demand stands out: a block on releasing self-improving models without demonstrated safety and alignment with human values.
OpenAI denies wrongdoing. Spokesperson Drew Pusateri said ChatGPT already has age-specific protections, parental controls and break reminders, and argued that the smarter path is industry-wide AI policy rather than lawsuits against individual companies.
The dollar amount isn't the point here. What matters is that the action is collective and bipartisan — nothing like it has happened to an AI company before. It lands a week after a report called ChatGPT "an unacceptable risk" for teenagers, and months after Florida filed the first lawsuit of this kind. If even part of the states' demands survive in court, it could set a bar for the whole industry, not just OpenAI.
