3 Oct 2026
NEWS · CONTEXT · PERSPECTIVE

News in perspective.
Context. Clarity. Sources.

Hot topics, news and trends shaping the world
Trending Israel's Supreme Court overturns ban on Arab parties in election G7 releases oil reserves amid diesel price crisis India's hockey gold and Olympic qualification AI cybersecurity scrutiny

California Attorney General Subpoenas OpenAI in Cybersecurity Probe

California is seeking more information from OpenAI about AI cybersecurity incidents and risks. The subpoena is part of an ongoing inquiry, separate from a federal FTC investigation; the supplied reports establish no final finding.

Compiled by issue.red from the sources below. Editorial policy

A sealed legal envelope and a laptop with abstract technology symbols, with a capitol building blurred in the background.
Conceptual illustration by issue.red; not a photograph of the event. Illustration: issue.red

A request for more information

California Attorney General Rob Bonta’s office says it has served OpenAI with an investigative subpoena in an ongoing inquiry into cybersecurity incidents and risks involving the company and its AI models. The California Department of Justice announcement and Reuters’ October 1 report describe this step. The investigation seeks information; the supplied accounts do not establish wrongdoing or legal liability. [1] [2]

An existing investigation, with questions still undisclosed

The California DOJ says it had already opened a formal investigation into the Hugging Face incident and was monitoring broader AI-industry compliance with state law. Bonta’s office says it is asking OpenAI additional questions about cybersecurity incidents and risks. The announcement supplied for this article does not set out the detailed questions in the subpoena or disclose a response deadline. [1]

How the attorney general frames the risks

Bonta’s announcement acknowledges that frontier AI models can serve legitimate cyber-defence purposes. It also states that developers have a responsibility to prevent their systems from perpetrating or enabling attacks, during development and testing as well as after release. These are positions expressed by the attorney general’s office. An investigative subpoena is not a court judgment, and neither supplied account contains a final finding in the California inquiry. [1] [2]

Separate state and federal proceedings

Reuters also reported a separate Federal Trade Commission investigation involving OpenAI and Anthropic, announced on Wednesday, September 30, concerning AI cybersecurity risks. That federal inquiry is distinct from California’s investigation. Reuters said OpenAI had not immediately responded to its request for comment. The supplied reports contain neither a substantive company response nor a final outcome for either investigation; that limit should not be read as a claim that no other public evidence exists. [2]

Sources

  1. As Part of Ongoing Investigation, Attorney General Bonta Serves Investigative Subpoena on OpenAI — California Department of Justice
  2. California attorney general issues investigative subpoena to OpenAI — Reuters

Corrections

  • Clarified that October 1 is the Reuters reporting date, rather than a confirmed subpoena service date. Removed an inference about the inquiry's scope and an unsupported blanket statement about public evidence. These errors had passed the evidence check.