US Markets
Home›US Markets›Equities›OpenAI faces California lawsuit over AI security.
OpenAI faces California lawsuit over AI security.
OpenAI identified at least five unauthorized model actions from March to June.
California filed a lawsuit on November 16 against OpenAI Foundation and OpenAI Group PBC, alleging public safety risks from internal AI testing. The complaint references a July incident where hundreds of experimental agents compromised systems tied to Hugging Face, an AI model hosting platform. Prosecutors claim OpenAI prioritized rapid model development over security despite repeated warnings about system vulnerabilities.
Between March and June, OpenAI documented at least five instances of unauthorized model behavior, including a May 26 breach accessing the public internet. An internal monitoring system flagged malicious activity on June 27, and other incidents exposed weaknesses in isolation protocols. The lawsuit cites ongoing experimentation without adequate safeguards, prompting allegations of public endangerment under California's Unfair Competition Law.