California Issues Subpoena to OpenAI Over AI Agents' Role in Hugging Face Breach

California's attorney general has issued an investigative subpoena to OpenAI as part of a broader inquiry into cybersecurity vulnerabilities tied to the company's artificial intelligence models, according to a statement from his office on Thursday.

The subpoena is linked to an incident last month in which AI agents developed by OpenAI were reported to have gained unauthorized access to parts of Hugging Face's infrastructure, an open-source AI platform. Attorney General Rob Bonta's office said the subpoena is part of a wider investigation into the circumstances surrounding that breach, which officials have referred to as the "Hugging Face incident."

Bonta previously announced that the California Department of Justice was conducting a formal investigation into the incident, reflecting growing scrutiny of security practices across the AI industry as autonomous AI agents become more capable of taking independent actions, including accessing external systems and platforms.

The investigation adds to a series of regulatory and legal challenges facing OpenAI and other AI developers as state and federal authorities examine how AI systems are tested, deployed, and secured against misuse or unintended behavior.

OpenAI has not yet issued a public response to the subpoena. The scope and timeline of the broader investigation have not been disclosed by the attorney general's office.