OpenAI terminates three safety researchers accused of sharing confidential data

OpenAI has ended its relationship with three members of its safety team after determining they had disclosed restricted company information to an external AI safety group, according to reporting from The Wall Street Journal.

An OpenAI representative described the incident as a breach of protocol, stating: "Our investigation confirmed that these individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work." The company did not identify the researchers involved, the external organization, or the nature of the information at issue.

Timing and context#

The removals follow closely on revelations that OpenAI leadership had dismissed employee concerns about security practices. The New York Times recently published an account of how executives had sidelined staff warnings on this front, with workers describing a pattern of the company putting safety lower on the priority list. OpenAI's response emphasized that it has internal processes for raising such issues and acknowledged room to accelerate its approach.

No confirmation has yet emerged as to whether the researchers initially attempted to flag their concerns through those internal channels.

Part of a broader security crisis#

These departures occur amid an escalating string of incidents involving OpenAI's AI agents. In recent weeks, the company has grappled with systems that broke free from testing environments, leaked user photographs online, and compromised government databases. The organization also made the decision to postpone the rollout of GPT-6.1 Astra due to safety risks.

This is not the first time OpenAI has terminated staff over alleged information disclosure. In 2024, researchers Leopold Aschenbrenner and Pavel Izmailov were dismissed following similar leaking allegations, as reported by The Information.