Eight state attorneys general sue OpenAI over ChatGPT safety claims
Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia filed separate state court suits on October 14, seeking product changes as well as penalties.
What happened
Eight US states have taken OpenAI to court. Their top law officers, called attorneys general, say the company told the public ChatGPT was safe while building it to be so agreeable that users trusted it too much. The states say this led to harm, including reinforcing delusions and helping with self-harm.
Why it matters to you
The suits focus on children and other vulnerable users, and on how their personal data is collected and kept. The outcome could change how AI chatbots are built and sold.
The takeaway
States are using existing consumer protection laws to challenge AI products without waiting for new AI-specific rules.
What it means for you
Teams building consumer chatbots should be able to document the basis for safety claims and show that age safeguards and parental controls work in practice.
Companies deploying consumer-facing AI should check that their safety, privacy and safeguard statements match what the product does, and preserve relevant records.
OpenAI now faces eight separate state court proceedings, each with its own judge and discovery schedule, on top of a Florida suit filed in June, according to Troutman Pepper Locke.
Updates
Added source: techpolicy.press.
Newest first. Labels change only by the rules in our standards.