California sues San Francisco's OpenAI over rogue AI agents
Add Axios as your preferred source to
see more of our stories on Google.

Illustration: Sarah Grillo/Axios. Stock: Getty Images
California Attorney General Rob Bonta is suing OpenAI, accusing the artificial intelligence giant of putting the public at risk by allowing experimental AI agents to escape its testing environment and launch cyberattacks.
Why it matters: The lawsuit marks a significant escalation in California's efforts to hold AI companies accountable for the potential dangers of an increasingly powerful technology that's raising questions about whether existing safeguards can keep pace with development.
Driving the news: Bonta filed the lawsuit Monday in San Francisco Superior Court, alleging OpenAI knowingly tested advanced AI models without adequate security measures and misled the public about its safety practices.
- "Without sufficient safeguards and controls, the scope and frequency of this already unacceptable risk will only increase," Bonta said at a press conference in San Francisco Monday morning.
Catch up quick: The complaint centers on a July incident in which hundreds of autonomous AI agents allegedly carried out a coordinated cyberattack against the open-source AI platform Hugging Face over more than two days.
- State prosecutors argue the rogue attack was preventable and reflected broader failures in how OpenAI developed and monitored its technology.
The big picture: The case puts one of San Francisco's most influential tech companies at the center of a growing debate over how aggressively governments should regulate AI as companies race to develop more capable systems.
State of play: OpenAI lost control of its experimental AI agents at least 80 times during testing, per Bonta.
- Prosecutors allege the incidents included unauthorized access to software databases, public information websites and government systems containing sensitive health information.
- Some agents also allegedly compromised OpenAI's own internal infrastructure, according to the lawsuit.
Between the lines: The lawsuit also accuses OpenAI of creating a public nuisance and violating California's Unfair Competition Law.
- Bonta is seeking court-ordered remedies, $2,500 per violation and the return of certain profits, Bloomberg Law reports.
What they're saying: OpenAI spokesperson Drew Pusateri told Axios that the company has published multiple reports, notified affected organizations and created more robust safeguards since the Hugging Face incident.
- "This lawsuit appears to single out OpenAI over a challenge that's been identified by every major AI lab," Pusateri said in an emailed statement.
- "We take misalignment seriously, and we have engaged — and will continue to engage — constructively with the Attorney General's office to share information and advance strong, practical safety standards for the entire industry."
What's next: The case could set an important precedent for how AI companies are held accountable and also help determine how far existing state laws can go in regulating the industry without new legislation.
