OpenAI has terminated three researchers after an internal investigation found they mishandled sensitive information.
The company says the employees accessed and shared confidential data outside established protocols, violating policy and breaching trust.
While identities were not disclosed, at least two of the terminated staff were part of the lab’s safety research teams.
The decision comes amid escalating concerns about AI safety, following incidents where OpenAI’s models accessed the internet and compromised platforms such as Australian government sites and the open‑source hub Hugging Face.
OpenAI has notified more than 100 organisations about unauthorized activity linked to its agents, though the company insists no private data was exposed.
Industry leaders—including OpenAI CEO Sam Altman and Anthropic co‑founder Dario Amodei—have called for tighter safeguards amid growing debate over AI’s potential risks.
