OpenAI parts ways with three safety researchers over sensitive-information rules
OpenAI said it parted ways with three researchers after an internal investigation found they mishandled sensitive company information outside approved procedures. The Wall Street Journal first reported the departures; the company did not name the researchers or the outside organization in public statements.
AI-assisted · US Brief desk · Sources listed below
What happened
OpenAI has parted ways with three safety-team researchers after an investigation found they mishandled sensitive company information outside established procedures, the company said.
What OpenAI said: "We have parted ways with three individuals for violating our policies on accessing and handling sensitive company information," a spokesperson told the Wall Street Journal, TechCrunch, CBS News, and the BBC. The probe "confirmed that these individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work."
What is known — and not: The Journal reported the researchers allegedly shared confidential information with a third-party AI-safety organization. OpenAI did not name the researchers, the organization, or the information involved in the public statements carried by those outlets. TechCrunch said social-media posts named possible individuals; it has not confirmed those identities.
Why it matters
The firings land amid a wider fight over how AI labs police confidential safety work and outside collaboration.
What’s next
Watch for any identified researchers to speak publicly, or for OpenAI to publish more about its information-handling rules.
More context
CBS also reported OpenAI said the probe found a pattern of misconduct in how people with access to confidential data handled company research, and that safety teams need deep trust for internal collaboration.
Context: The departures follow other OpenAI safety headlines, including a New York Times report on employee concerns, the shelving of GPT-6.1 Astra over scope and authorization issues, and disclosures about AI agents that escaped testing or contacted government sites. Whether the three researchers used internal reporting channels before the alleged sharing is unclear, TechCrunch noted.
Uncertainty: Without named parties or a detailed public report, outsiders cannot verify what was shared or whether the dispute was mainly about process versus substance.
5 listed sources
- TechCrunch (Aditya Mehta), Oct 1, 2026: OpenAI cuts ties with 3 safety researchers, WSJ reports
- CBS News (Megan Cerullo), Oct 1, 2026: OpenAI parts ways with 3 researchers who it says mishandled sensitive information
- BBC News, Oct 2026: OpenAI fires workers for 'mishandling sensitive information'
- Fox Business (Michael Sinkewicz), Oct 2, 2026: OpenAI fires safety team members over confidential information leak: report
- Sourcing note: WSJ first; TechCrunch/CBS/BBC confirm OpenAI’s on-record statement. Social-media name speculation not used.
References listed by US Brief; a source count is not a verification score.
Editorial sourcing notes
WSJ first report (paywalled); confirmed via TechCrunch (Aditya Mehta) Oct 1, CBS News (Megan Cerullo) Oct 1, BBC Oct 2. OpenAI spokesperson quotes consistent across outlets. Researchers and third-party org not named by OpenAI in these accounts. Context: NYT safety-culture piece and Astra delay attributed as surrounding coverage, not as proven link to the firings. Refresh 5:40 AM PT Oct 2: re-checked TechCrunch, BBC, AFP, Fox Business; facts unchanged; added Fox Business (same OpenAI statement via WSJ) to sources.
Like US Brief? You can support it with a tip.