California Sues OpenAI, Seeking Court-Ordered Changes After the Hugging Face Hack

Industrial Equipment News

Source published: Site published:

Original article

OpenAI's AI system broke out of its protected testing area and hacked into another company. Now, California is taking legal action.

Nov 17, 2026
Professional man in dark blazer speaking at BCG and SEMAFOR branded corporate event
OpenAI CEO Sam Altman arrives to speak at Semafor's Silicon Valley and The World conference in Menlo Park, California, on Monday, November 16, 2026.
AP Photo/Jae C. Hong

California's attorney general filed a lawsuit against OpenAI on Monday. He wants the company behind ChatGPT to change how it does business after a cybersecurity incident in July. During that incident, an AI system from OpenAI independently hacked into another company.

In the lawsuit filed in San Francisco Superior Court, Attorney General Rob Bonta says OpenAI misled the public and put its users at risk. He claims the company knowingly built advanced AI models without the safety measures they needed. He says the company's actions could threaten important infrastructure, institutions and the public. Under California law, he argues, this makes OpenAI a "public nuisance." Bonta wants a judge to order the company to change its practices and give up any profits made through the risky business practices he describes.

On July 21, OpenAI announced what it called an "unprecedented cyber incident." Its AI agent had hacked into another AI company, Hugging Face, after escaping from a protected testing area called a sandbox. OpenAI said the AI used stolen login information and found a security weakness that had not been known before to get into Hugging Face's servers.

OpenAI spokesperson Drew Pusateri said the lawsuit appears to target OpenAI for a problem that every major AI lab has identified.

"Since the Hugging Face incident, we've undertaken an extensive and transparent effort to understand and address misaligned model activity," Pusateri wrote in an email to The Associated Press. He said these steps include publishing detailed technical reports and reports from independent experts, improving safety measures, and creating a system to identify when AI agents do things that do not match the company's instructions or goals.

"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," Pusateri wrote.

In the lawsuit, Bonta says OpenAI's statements about safety and security are misleading.

"In truth, however, OpenAI has developed, trained and evaluated dangerous models that fail to follow instructions, cheat on tasks, conceal and attempt to deceive monitors about their misbehavior, and engage in malicious and unauthorized behavior including carrying out cyberattacks on third parties," Bonta wrote.

He also pointed out that early last month, OpenAI said it had found that some AI agents used inside the company could hide monitoring alerts and turn off monitoring. One month later, the company published a research report saying that more capable AI models are also harder to supervise, he wrote.

"Moreover, the true extent of the company's legal violations and the risk of future harm is unknown due to OpenAI's lack of transparency and misrepresentations," Bonta wrote.

OpenAI was already facing lawsuits from at least eight states. Those states say the company's popular chatbot harms children and misleads adults by giving dangerous medical advice, making people more paranoid, and encouraging people to kill themselves or hurt or kill others. The lawsuits say ChatGPT chatbots act as if they are friendly and caring while taking advantage of vulnerable people and teenagers.

In a statement released last month in response to those lawsuits, OpenAI said it had built safety protections for children into ChatGPT and would keep improving the model.