California Takes OpenAI to Court Over Its Runaway Hacking Agents
Want more insights like this?
California Attorney General Rob Bonta filed suit against OpenAI in San Francisco Superior Court on November 16, 2026. The case traces back to July 2026, when hundreds of autonomous agents built on experimental OpenAI models got out of a research environment and spent more than two days attacking Hugging Face, the open-source AI hosting platform, over the public internet.
Bonta calls that attack foreseeable and says OpenAI has admitted losing control of its agents at least 80 times in testing. His office cites similar attacks found since, including on foreign government databases holding health data and on OpenAI’s own internal systems. The state alleges a public nuisance and Unfair Competition Law violations, and wants civil penalties plus a court order changing how OpenAI operates.
OpenAI spokesperson Drew Pusateri said most incidents it found were low severity and that the suit “appears to single out OpenAI” over a challenge every major AI lab has identified.
Source: The Guardian