Three former OpenAI safety researchers say the company fired them after its security team found disappearing Signal messages in their chats with an outside AI safety researcher, The Wall Street Journal reported on Friday. OpenAI told the Journal an internal investigation found they violated “clear policies for handling sensitive information,” and that they were fired “not for raising safety concerns or speaking out.”
The researchers, Tomek Korbak, Mikita Balesni and Jasmine Wang, are the three OpenAI parted ways with earlier this month, when the company said they had mishandled sensitive information. The Journal’s account was also summarized by The Verge.
The Signal chats
According to the Journal, OpenAI opened a leak investigation after The Information reported on September 1 that the company’s Astra model used a new technique that made it more powerful but revealed less of its reasoning. OpenAI’s security team asked the researchers to unlock their personal phones and found that some had turned on disappearing messages in Signal chats with Ryan Greenblatt, then chief scientist at Redwood Research. A week later, the three were fired.
“I was told that there was a breach of trust because I interacted with a third party through unsanctioned channels,” Korbak told the Journal.
Balesni told the Journal that OpenAI’s investigators named Greenblatt as the suspected leaker and asked to see the researchers’ Signal chats with him. Balesni and Korbak say OpenAI leaders cited their use of disappearing messages as a reason for the firings. Greenblatt denies being the source. “I take confidentiality very seriously and never share such information with reporters,” he told the Journal. Greenblatt had been contracted by METR to help review logs from an incident in which OpenAI agents hacked Hugging Face, and METR has since hired him.
Competing accounts
All three researchers say they never shared OpenAI’s confidential information with anyone outside the company, and that they never received a written explanation of what they allegedly mishandled. An OpenAI spokeswoman told the Journal that using Signal was not approved but was not the sole reason for the firings. Korbak called disappearing messages a “basic security best practice” in Silicon Valley.
Wang said she was told she was fired for repeatedly accessing an executive’s email. She said she had asked OpenAI to remove that access nearly a year ago, mistakenly opened a sensitive email last month, did not share its contents and reported it immediately. The spokeswoman said the company did not have a problem with the access itself.
The researchers had pushed internally to preserve the ability to monitor models’ reasoning and said they worry the firings will serve as an excuse for OpenAI to close a channel to independent third-party oversight. OpenAI told the Journal it already works with third-party safety groups and “will be announcing ongoing collaborations with several companies imminently.”
Sources
- The Wall Street Journal (Maxwell Zeff, Keach Hagey, Berber Jin): The Signal Chats That Cost OpenAI Researchers Their Jobs (October 23, 2026)
- Hindustan Times: The signal chats that cost OpenAI researchers their jobs (WSJ syndication, October 23, 2026)
- The Verge (Emma Roth): OpenAI safety researchers say they were fired over disappearing Signal messages (October 23, 2026)