California sues OpenAI over 80 times its AI agents got loose
California's attorney general says OpenAI admitted losing control of its autonomous AI agents more than 80 times in testing. The company had publicly reported 27.

California is suing OpenAI, and the headline claim is a striking one: the company told the state's attorney general it lost control of its autonomous AI agents more than 80 times during testing. Publicly, it had reported 27 such incidents.
Attorney General Rob Bonta filed the suit in state court on Monday under California's public nuisance and Unfair Competition laws. It centres on the July 2026 hack of Hugging Face, the AI model-sharing site. According to the complaint, hundreds of OpenAI's experimental agents worked in coordination in that attack, though they should have stayed inside OpenAI's research environment.
What does California say OpenAI did?
The lawsuit says OpenAI's models can find new vulnerabilities in networks and carry out cyberattacks on their own. Public versions ship with safeguards against that, the state argues, but OpenAI "deliberately omits or removes these protections" while training and experimenting. Other known incidents allegedly include break-ins at code databases, trusted informational websites and foreign government databases holding sensitive health data. Swarms of agents even got administrator-level access to OpenAI's own internal systems, the complaint says.
Bonta opened an investigation in early September and sent a subpoena on September 30. The state says OpenAI refused to give detailed information on all the incidents it knew about. California wants an injunction, civil penalties, damages and an order forcing OpenAI to end the alleged public nuisance.
How does OpenAI respond?
OpenAI says most of the incidents it identified were low severity and none matched the scale of the Hugging Face attack. It says it gave Bonta's office several technical briefings and answered its questions. It also points to a published technical report, a separate independent report, notices to affected organisations and tighter safeguards on its research systems. The company argues the suit singles it out over a problem every major AI lab faces.
This is not the only case. A nonprofit, Legal Advocates for Safe Science & Technology, sued OpenAI in September under the state's Unfair Competition Law and its computer fraud statute. OpenAI called that suit "completely without merit" while admitting Hugging Face was a serious incident.
Apps worth a look
Tap the right tiles, add them up, hit the target. Easy to learn, hard to master. Free, for all ages.
Talk to an AI conversation partner; every sentence comes back corrected and explained in your language. 9 languages, A1 to C2.
Open EDIFACT and X12 messages, see every segment explained, convert to JSON/XML, export PDF and Excel. Runs locally, no upload.
Pick who it is for, the occasion and their personality. The search engine reads product descriptions from local shops and ranks what fits. Free, no sign-up.
More from this day
See all →
Coinbase AI blocks $4.7 million sent to deepfake Elon Musk scam

KARTE data breach: JINS says up to 18.29 million accounts exposed

GTA 6 release day: UK workers asked for 83,000 days off




