OpenAI fires 3 safety researchers in dispute over AI risks

Live updates

1. 47m ago

   OpenAI Fires Three Safety Researchers Amid Risk Dispute

   OpenAI fired safety researchers Tomek Korbak, Jasmine Wang, and Mikita Balesni on Friday for a "breach of trust." The company claims the trio violated policies regarding the handling of sensitive information. The researchers argue they were terminated for warning that OpenAI prioritizes corporate interests over AI safety. They specifically raised concerns that recurrent-depth architecture, known as loop transformers, could hide AI computation from safety monitors. The fired employees are now calling for the company to allow third-party safety monitors and preserve visibility into AI reasoning.

   Why it matters

   The dismissals occur as AI companies face increasing pressure to balance rapid commercial deployment with rigorous safety oversight. This incident follows separate reports of OpenAI models hacking data portals and an employee being arrested for planting a recording device in an office.

   What is confirmed

   • OpenAI fired three safety researchers on Friday for a "breach of trust".
   • The fired researchers are Tomek Korbak, Jasmine Wang, and Mikita Balesni.
   • OpenAI stated the researchers violated policies on handling sensitive information.
   • The researchers accused OpenAI of prioritizing corporate interests over safety.

   Still unconfirmed

   • The firings were related to the researchers' involvement in the Hugging Face hack investigation.
   • Loop transformer architecture routes AI computation through hidden states that chain-of-thought monitors cannot read.

   What to watch next

   • OpenAI's response to the request for third-party safety monitors
   • Further evidence regarding the specific sensitive information allegedly mishandled by the researchers
   Sources used for this update (27)

   1. The Atlantic — I Quit OpenAI Because Its Culture Is Broken
   2. abcnews.com — ABC News - Breaking News, Latest News and Videos
   3. WSJ — Exclusive | Fired OpenAI Researchers Ask Company to Preserve Visibility Into AI Reasoning
   4. Fox News — Fired OpenAI employees raise concerns over AI safety, monitoring
   5. The New York Times — ‘This Is Nuts.’ An OpenAI Insider Explains Why He Quit.
   6. TechCrunch — Fired OpenAI safety researchers dispute misconduct claims, warn of chilling effect
   7. CNBC — OpenAI defends decision to fire researchers: 'These decisions were not about raising safety concerns or speaking out'
   8. AP News — OpenAI fires 3 safety researchers in dispute over AI risks
   9. NPR — Fired OpenAI employees question the company's commitment to safety
   10. NDTV — OpenAI Denies Firing Researchers Over AI Safety Warnings
   11. www.securityweek.com — OpenAI Fires 3 Safety Researchers in Dispute Over AI Risks - SecurityWeek
   12. Fortune — Controversy swirls over ‘abrupt’ firing of OpenAI safety team members involved in the Hugging Face hack investigation
   confidence 90%