OpenAI parts ways with three safety-team staff over sensitive-information policies
OpenAI said Thursday it has parted ways with three individuals for violating its policies on accessing and handling sensitive company information, after the Wall Street Journal reported the exits involved safety-team researchers allegedly sharing confidential details with a third-party AI-safety organization.
OpenAI says it cut three people for mishandling confidential data, while the Journal reports they were safety-team staff who allegedly shared that work with an outside safety group. Until the company says what was shared and with whom, the story is as much about who gets to test the most capable models as it is about three jobs.
On Thursday, 1 October 2026, the Wall Street Journal reported that OpenAI had parted ways with three researchers for allegedly sharing confidential company information with a third-party AI-safety organization. The story is by Keach Hagey, Maxwell Zeff, and Berber Jin. The page stamps Oct. 1, 2026, 12:20 p.m. Eastern. The Journal says that account comes from people familiar with the matter. It says the company recently told some employees it had terminated three researchers who worked on its safety team, and it attributes that sentence to one of those people. A third party, here, is a group that is not OpenAI. Those lines are the Journal’s. OpenAI did not post them on its own site.
What OpenAI said on the record is a policy finding, not a list of the materials. A spokesperson told the Journal, in the statement TechCrunch quotes: “We have parted ways with three individuals for violating our policies on accessing and handling sensitive company information. Our investigation confirmed that these individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work.” Three is the company’s count of people. Sensitive company information is information the company treats as confidential. Established procedures are the company’s rules for who may see that information and how it may leave. The statement says “individuals.” It does not say “researcher,” and it does not say “safety team.” It does not name the people, an outside group, or the materials. Those lines are OpenAI’s, as the Journal was told them.
CBS News carried the investigation sentence as a statement to CBS. Megan Cerullo’s item, updated October 1, 2026, at 2:04 p.m. Eastern, says OpenAI has parted ways with three employees who the company says violated its policies for handling sensitive company information. CBS says the company added that its probe uncovered a pattern of misconduct in how individuals with access to confidential data handled company research. A pattern, in that sentence, means the company says the conduct showed up more than once. The spokesperson added that OpenAI’s safety teams see internal insights that require deep trust, and that without that trust people inside the company cannot work together. CBS says the company did not specify what kind of information was mishandled, or how. CBS also says the Journal was first to report that three OpenAI researchers had departed, allegedly for sharing confidential information with a third-party AI-safety organization. The safety-team setting and the outside organization are the Journal’s account, which CBS relays. The pattern line and the trust line are CBS’s account of what the company added.
TechCrunch, by Aditya Mehta, stamps 11:14 a.m. Pacific on October 1, which is 2:14 p.m. Eastern. It says the Journal reported three researchers on the safety team who allegedly shared confidential company information with a third-party AI-safety organization. It quotes the same two sentences as the statement to the Journal. It says the report did not name the researchers, the organization, or the information involved, and that OpenAI did not immediately answer TechCrunch. It says posts on X named people some users believe were among those dismissed, and that TechCrunch has not confirmed those identities. An unconfirmed social post is not an identification by OpenAI or by the Journal.
The cited pages do not name the three people, the outside organization, or the materials. They do not describe the exits as legally protected whistleblowing. Whistleblowing, in that use, would mean a disclosure the law protects. That status is not in these reports. The company’s confirmed facts are the parting, the policy violation, and the finding that sensitive information was handled outside its procedures. The safety-team jobs and the alleged share with an outside AI-safety group stay with the Journal’s people familiar with the matter.
The picture is a desk graphic. A navy panel reads OpenAI, then a large 3, then “safety-team” and “staff parted,” and a line that the company statement is dated Oct. 1, 2026. A brick-colored rule splits the frame. The cream side’s headline reads “Parts ways with three over sensitive-info policies.” The bullets say OpenAI says three individuals violated policies on accessing and handling sensitive company information; the investigation found mishandled data outside established procedures; the Journal alleged sharing with a third-party AI-safety organization that is unnamed; and the names, the organization, and the materials are not disclosed. The footer reads “WSJ exclusive + OpenAI on-record statement.” The graphic summarizes those two accounts. It is not a photograph of the three people, and it does not name them.
In plain terms, OpenAI said on Thursday that it parted ways with three people after an investigation found they mishandled sensitive information outside company procedures. The Wall Street Journal, first on the exits, reported that the three worked on the safety team and allegedly shared confidential information with a third-party AI-safety organization. The company and the Journal have not named the people, the organization, or what was shared.