Arkansas sues OpenAI, alleging it misled users about ChatGPT safety

Arkansas Attorney General Tim Griffin has sued OpenAI, saying the company told users ChatGPT was safe while it knew of serious risks.

The 73-page complaint was filed on Wednesday in a Faulkner County court and names eight OpenAI firms.

The state says ChatGPT has encouraged suicide and self-harm, fed delusions, given bad medical advice and helped people plan violence. It says OpenAI knew about these risks yet still sold the chatbot as safe and suitable for teens.

The suit uses two state laws: the Arkansas Deceptive Trade Practices Act and the Children and Teens’ Online Privacy Protection Act. Under the second, the state says OpenAI collects children’s data without proper age checks and retains it for too long.

The case also goes beyond the chatbot. It cites cases this year in which OpenAI’s AI agents broke out of their test setups. In one case, in July, agents running a security test gained access to Hugging Face’s servers.

“OpenAI represented to Arkansans that ChatGPT was safe, but they knew ChatGPT had significant risks and limitations,” Griffin said in a statement.

The state wants fines of $10,000 per violation, plus another $10,000 for each one that hurt an older or disabled person. It also wants money paid back to users, and damages.

In addition, Arkansas wants the court to stop OpenAI testing its most advanced models until outside monitors confirm its safeguards work.

It also wants a ban on models that can improve themselves until OpenAI can show they will act in line with human values.

Arkansas was one of eight states that sued OpenAI on Wednesday, Platformer reported . The others were Arizona, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia, each in its own state court.

Florida went first , suing OpenAI and its chief executive, Sam Altman, in June. Then, in September, it asked a court to halt ChatGPT development while the case runs, a request similar to Arkansas’s.

In a statement on Wednesday, OpenAI spokesperson Drew Pusateri said ChatGPT has age-specific protections. He added that the company will continue to share details of cases in which its AI acted against instructions.

“While no system can be perfect, we’ll continue to work with experts to inform our work and build additional safeguards, such as the parental controls, break reminders and age-appropriate model protections that exist today,” Pusateri said.

None of the claims have been tested in court, and Griffin has asked for a jury trial.