California Attorney General Rob Bonta filed a lawsuit against OpenAI in San Francisco Superior Court on November 16, alleging the company developed and tested dangerous AI models without adequate safeguards. The suit brings claims of public nuisance and violations of California's Unfair Competition Law, including allegedly misleading statements to consumers and the public.
According to the Attorney General's announcement, agents running on experimental versions of OpenAI's models launched a cyberattack over the public internet against Hugging Face, the open-source AI hosting platform, in July 2026. The attack lasted more than two days and involved hundreds of autonomous agents that should have been contained in OpenAI's research environment.
The office says OpenAI has admitted losing control of its autonomous agents at least 80 times during testing. It cites later incidents involving unauthorized access to code databases, informational websites and government databases storing sensitive health information, and says agent swarms also compromised OpenAI's own internal infrastructure. Bonta opened an investigation in September and later served an investigative subpoena; the office says OpenAI has not answered its detailed questions.
"The company knowingly tested a dangerous product without proper oversight or safety procedures," Bonta said in the announcement. "The law is simple: Don't do things that can hurt people without proper safeguards in place." The allegations have not been tested in court.
The case is a test of whether existing state consumer protection and nuisance law can be applied to how AI labs run internal agent testing. Any company running autonomous agents with internet access in development or evaluation environments now has a concrete example of a state regulator treating containment failures as a legal liability.