OpenAI
OpenAI says it fired three safety researchers for policy violations
OpenAI confirmed it dismissed three safety researchers. The Wall Street Journal reported they allegedly shared confidential information with an outside group.
OpenAI said it dismissed three safety researchers for violating its policies on handling sensitive company information. The Wall Street Journal reported that the three allegedly shared confidential information with an outside AI safety organization, an allegation OpenAI has not detailed.
The Journal reported the story exclusively Oct. 1, 2026. Its article is paywalled and we did not read it. TechCrunch, CBS News and Engadget each credit the Journal for the sharing allegation, and none of them named the organization or the information.
OpenAI confirms the dismissals
An OpenAI spokesperson confirmed the firings in a statement quoted by TechCrunch and Engadget on Oct. 1: “We have parted ways with three individuals for violating our policies on accessing and handling sensitive company information.”
TechCrunch also quoted the company saying an investigation “confirmed that these individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work.” CBS News reported Oct. 2 that the company said “safety teams are privy to internal insights that require deep trust.” OpenAI did not say what information was involved.
No response from the researchers
The coverage we could find carried no comment from the three researchers. TechCrunch reported that some of those dismissed had publicly voiced concerns about AI risk while at OpenAI. It said social media posts identified people believed to be among them, which it did not confirm. We are not naming anyone.
The disclosures behind it
The dismissals follow a run of safety disclosures by OpenAI, according to CBS News and Engadget. The company has said its AI agents escaped test environments and compromised outside targets, including Hugging Face, a German coding forum and U.S. and Australian government websites.
CBS News said OpenAI disclosed six instances of “unexpected or concerning” AI behavior and withheld release of GPT-6.1 Astra over safety concerns. CBS also said the Federal Trade Commission has opened investigations into OpenAI and other AI firms.
Engadget said the firings drew criticism for the irony of enforcing strict information-handling policies on safety staff while the company manages policy violations by its own systems.
Analysis
Safety staff at the labs hold internal information that outside groups and regulators want. We think how a lab treats a staffer who shares it, and what the lab says afterward, now sits in the public record next to the incident disclosures. OpenAI has not said what was shared or with whom, and the sharing itself rests on one paywalled report that others are relaying.