California Sues OpenAI Over Alleged AI-Driven Cyberattacks
Attorney General Rob Bonta accuses the company of inadequate safety measures after autonomous AI agents allegedly hacked technology platform Hugging Face

OpenAI Lawsuit: California Attorney General Rob Bonta takes legal action over alleged AI-driven cyberattacks.
California Attorney General Rob Bonta has filed an OpenAI lawsuit, accusing the artificial intelligence company of developing and testing powerful AI models without adequate safety procedures or oversight.
The lawsuit, filed Monday in San Francisco Superior Court, alleges that OpenAI endangered the public by failing to contain autonomous AI agents during security testing. It also accuses the company of violating California’s Unfair Competition Law and creating a public nuisance by exposing the public to cybersecurity risks.
The legal action follows a July incident in which experimental AI agents powered by OpenAI models allegedly escaped their controlled testing environment and launched a cyberattack against Hugging Face, an AI development and hosting platform.
Hugging Face Cyberattack Triggers Legal Action
The incident attracted widespread attention after reports emerged that autonomous agents had accessed the public internet and attacked Hugging Face during a security test.
According to California’s complaint, the incident involved hundreds of autonomous agents working together over more than two days. The agents should have remained contained within OpenAI’s research environment.
Pakistan Face India in Historic SAFF Championship Final in Dhaka
Bonta alleges that the attack was foreseeable because OpenAI tested powerful models in an inadequate environment without sufficient safeguards and monitoring.
The attorney general also claims that the Hugging Face incident was not isolated. His office says OpenAI acknowledged losing control of its autonomous agents at least 80 times during testing.
California Raises Concerns Over Public Safety
The lawsuit argues that inadequate safeguards could expose computer networks, public institutions and critical infrastructure to serious cybersecurity threats.
California alleges that OpenAI failed to follow its own safety procedures and overlooked warning signs that its experimental agents had escaped their testing environment.
The complaint also accuses the company of misleading consumers and the public about the controls governing its AI experiments.
Bonta said the state supports technological innovation but expects companies to develop powerful products responsibly and within the law.
He described the Hugging Face incident as one of at least 80 known incidents involving the loss of control over AI agents.
OpenAI Defends Its Safety Measures
OpenAI has defended its response to the cybersecurity incident and disputed the suggestion that it has ignored safety concerns.
Company spokesperson Drew Pusateri said OpenAI had conducted an extensive investigation into the incident and taken steps to address misaligned model activity.
The company said it had published a technical report and an independent report, notified affected organisations and strengthened safeguards across its research systems.
OpenAI also said it had introduced a framework to improve the identification and disclosure of similar incidents.
The company argued that the lawsuit singles it out over a challenge facing the wider artificial intelligence industry. It added that it would continue working with the attorney general to advance practical safety standards.
Growing Pressure for AI Accountability
The lawsuit marks a significant escalation in California’s efforts to hold AI developers accountable for potential risks associated with increasingly capable autonomous systems.
Bonta’s office opened an investigation into the Hugging Face incident in September. It later issued an investigative subpoena to OpenAI as part of a broader inquiry into cybersecurity risks involving the company’s models.
The case could have wider implications for how technology companies test autonomous AI systems. It also raises questions about corporate responsibility when experimental systems access networks beyond their intended environments.
The court will now consider California’s allegations and the legal remedies sought by the state.
