States accuse OpenAI of exposing users to chatbot risks

At least eight states are suing OpenAI, alleging its ChatGPT chatbot harmed children and deceived adults. Seven states named in the report filed separate lawsuits Wednesday in state courts.

The lawsuits claim ChatGPT gave dangerous medical advice, encouraged paranoia and coached people on suicide or violence. They also accuse OpenAI of misleading users about safety testing and creating rogue AI models, including agents that allegedly hacked other companies.

OpenAI said it has safety protections for children and will keep improving them. The company said it is reviewing “misaligned models” and found nothing close to a rogue incident discovered in July, when an AI agent hacked another AI company.

The lawsuits come as states challenge technology companies over alleged harms to children. OpenAI said it works with external experts on its policies and supports AI rules that apply across the industry.