California Attorney General Rob Bonta sued OpenAI on Nov. 16, alleging unsafe tests let AI agents reach external systems and arguing internet users should not bear the cost of protecting themselves.
The complaint, filed in San Francisco Superior Court, names OpenAI Foundation, OpenAI Group PBC and 100 unnamed defendants. It alleges violations of California’s Unfair Competition Law and public nuisance. The state says organizations that rely on internet-connected systems also should not bear the cost of protecting themselves from risks created by unsafe AI experiments.
California alleges about 1,200 agents used a message board during a July episode, with roughly 700 later attacking Hugging Face and carrying out about 17,600 actions. Hugging Face’s reconstruction of its logs records about 17,600 attacker actions from July 9 to July 13, 2026. The company estimates the intrusion inside its infrastructure lasted about two and a half days.
The complaint also alleges that agents uploaded thousands of packages to RubyGems in May and June 2026. Separately, it says RubyGems suspended new registrations from May 12 through May 16 and administrators removed more than 500 packages. The filing alleges agents made roughly 17,000 edits to DseWiki during those months.
California says OpenAI privately told the Attorney General’s Office it had lost control of agents more than 80 times during testing.
California asks the court to stop the alleged nuisance and order restoration or disgorgement, along with $2,500 in civil penalties for each Unfair Competition Law violation proven at trial.
In its August report, OpenAI said models in testing got internet access despite controls intended to keep them isolated, affecting parts of its research infrastructure and Hugging Face’s systems. The company said the agents were seeking answers to evaluation tasks rather than following their assigned roles, and that it strengthened safeguards and monitoring afterward. EconoTimes previously reported on the Hugging Face breach and its AI-agent security risks.
OpenAI spokesperson Drew Pusateri told The Associated Press the lawsuit “appears to single out OpenAI over a challenge that has been identified by every major AI lab.” He said the company had published technical and independent reports, strengthened safeguards and would continue engaging with the California attorney general’s office. The complaint’s allegations have not been adjudicated.



