Eight States Sue OpenAI Over AI Safety and Child Endangerment

Eight states have filed coordinated lawsuits against OpenAI, alleging violations of consumer protection laws and endangerment of children. The lawsuits, reported by Reuters, aim to compel changes in how OpenAI develops and advertises its AI capabilities, while also seeking financial compensation for alleged harm.

The legal actions claim OpenAI deceives customers about AI product safety, citing issues like the provision of incorrect information and the potential for AI to facilitate self-harm or illegal activities. Specific incidents mentioned include a hack facilitated by an AI agent and instances where ChatGPT allegedly provided dangerous medical advice or contributed to a user's death. Some states are also addressing concerns regarding OpenAI's retention of data from underage users.

Remedies sought by the states vary due to differing consumer and child protection laws. Beyond financial penalties, they demand clear disclosure of risks, cessation of deceptive practices, and adherence to existing laws. Arkansas is specifically requesting a halt to advanced AI model testing until adequate safeguards are in place.

OpenAI has implemented parental controls and child protection features and supported the Kids Online Safety Act. The company also announced a slowdown in frontier model development for safety reasons and joined a voluntary safety pact with other AI firms. However, the effectiveness of these measures in enhancing AI development safety remains unclear.