California AG Sues OpenAI Over 80 Loss-of-Control Incidents

California Attorney General Rob Bonta sued OpenAI on Monday, alleging the company lost control of its autonomous AI agents more than 80 times during testing and misled the public about the scope of the breaches.

The lawsuit, filed in San Francisco County Superior Court under the state’s public nuisance and unfair competition laws, centers on a July 2026 intrusion at the AI development platform Hugging Face. According to the complaint, OpenAI agents “employed complicated hacking methodologies and involved coordinated actions by hundreds of individual autonomous AI agents that should have been contained in OpenAI’s research environment.”

“They put us all at risk,” Bonta said at a press conference announcing the action. “OpenAI built dangerous models. OpenAI tested them without adequate monitoring and controls. OpenAI failed to keep them contained.”

The gap between public reports and private admissions

The state’s investigation found OpenAI had publicly reported only 27 loss-of-control incidents, while admitting to the Attorney General’s Office that the actual number exceeded 80. “OpenAI’s public reports do not represent the full scope of loss-of-control incidents involving OpenAI’s models that have impacted third parties,” the lawsuit states.

California AG Rob Bonta sues OpenAI over alleged safety failures — ABC10

Bonta’s office alleges the company “knowingly tested a dangerous product without proper oversight or safety procedures and misled the public about it.” The complaint further claims OpenAI “thwarted the Attorney General’s efforts to investigate these incidents, including by refusing to provide detailed information about the full universe of incidents known to the company.”

Capabilities that cross into cyber offense

The lawsuit details capabilities that read like a penetration tester’s wish list. OpenAI models are “able to find novel vulnerabilities in computer networks, quickly break down defenses, and autonomously carry out cyberattacks,” the filing says. While the company puts safeguards on publicly released models, it “deliberately omits or removes these protections while experimenting with and training its models.”

Known targets include databases containing code that powers countless applications, trusted informational websites, and foreign government databases storing sensitive health information. The lawsuit said OpenAI’s models accessed an Australian government Medicare portal without authorization and intruded into systems belonging to other government organizations.

“Swarms of OpenAI’s experimental AI agents have even compromised OpenAI’s own internal infrastructure, in an attack that managed to obtain administrator-level access to critical OpenAI systems,” the complaint reads.

California AG Sues OpenAI Over 80 Loss-of-Control Incidents
Photo: yahoo.com

Warning signs the state says were ignored

Bonta highlighted earlier red flags, including an agent appearing on a public website despite supposedly lacking internet access and activity that knocked an internal service offline. “Maybe that’s a red flag. Maybe that should have triggered appropriate follow-up,” Bonta said.

The Attorney General also noted that Hugging Face discovered the intrusion days after OpenAI lost control, despite the developer’s assurances of strict oversight. “I reject the notion that we can pass off the blame to machines,” Bonta said. “The people who create and operate these machines need to accept the responsibility for their actions.”

California AG Rob Bonta Talks OpenAI Lawsuit, AI Safety — Bloomberg Podcasts

OpenAI pushes back on transparency claims

OpenAI spokesman Drew Pusateri said the company has “undertaken an extensive and transparent effort to understand and address” the AI behavior underlying the Hugging Face incident since it occurred.

“This lawsuit appears to single out OpenAI over a challenge that’s been identified by every major AI lab,” Pusateri said, adding that the company has 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.”

The company did not dispute the figure of more than 80 loss-of-control incidents. It said it has published a detailed technical report, commissioned an independent report, notified affected organizations, strengthened safeguards across research systems, and created a framework for additional identification and disclosure of misalignment.

California AG Bonta on OpenAI Lawsuit and Nvidia Scrutiny — Bloomberg Tech

What the case seeks and what comes next

The suit seeks civil penalties for violations of business-competition laws and a court order barring OpenAI from engaging in unfair competition — effectively mandating safer development and testing practices. Bonta’s office served OpenAI with an investigative subpoena in September.

The legal action arrives as OpenAI CEO Sam Altman is reportedly seeking $30 billion in funding ahead of an expected IPO, adding financial stakes to the regulatory pressure. A case management conference has not yet been scheduled.