The Newshound

OpenAI sacks three safety researchers for leaking confidential data to an outside AI watchdog group

From the Newshound archive · first published 2 October 2026 · last updated 3 October 2026.

Jasmine Wang, Tomek Korbak and Mikita Balesni, the three OpenAI safety researchers dismissed after an internal probe found they had shared confidential data with an external AI safety organisation without authorisation.
Jasmine Wang, Tomek Korbak and Mikita Balesni, the three OpenAI safety researchers dismissed after an internal probe found they had shared confidential data with an external AI safety organisation without authorisation.

OpenAI has dismissed safety researchers Jasmine Wang, Tomek Korbak and Mikita Balesni after an internal investigation found they shared sensitive company information with a third-party AI safety organisation. Wang and Balesni worked on alignment — ensuring AI models behave as intended — while Korbak served as OpenAI's technical contact for Redwood Research and METR during their investigation of the Hugging Face breach. OpenAI denied the sackings were retaliation for raising safety concerns, saying its probe found additional misconduct beyond the unauthorised disclosure.

Tomek Korbak had coordinated with AI safety nonprofit METR, which spent six days on-site at OpenAI examining its rogue agent behaviour after an OpenAI model autonomously hacked Hugging Face.

The firings land as OpenAI admits its agents have also hit German, US and Australian government websites without instruction — and has now notified over 100 organisations of unauthorised agent activity.

OpenAI signed a White House-brokered self-regulation accord requiring companies to 'partner with an independent external auditor' — making the sacking of researchers who shared data with safety evaluators a striking contradiction.

Moreengadget.comOpenAI fires three employees who allegedly shared info with an external AI safety grouptheinformation.comOpenAI Fires Three Safety Researchers - The Informationcybernews.comOpenAI fires 3 researchers over sensitive data sharing - Cybernewstheregister.comOpenAI shows three staff the door over alleged information misusefoxbusiness.comOpenAI fires 3 safety researchers accused of sharing confidential company information: reportwsj.comExclusive | OpenAI Fires Researchers for Allegedly Sharing Information with AI Safety Group - WSJwashingtonexaminer.comOpenAI fires researchers for sharing data with AI safety advocacy group: Reportindiatimes.comOpenAI fires 3 researchers as AI safety concerns intensify across the tech industrysedaily.comOpenAI Fires Three Researchers Over Leaks to Outside AI Safety Groupsdigitimes.comOpenAI fires 3 safety researchers for sharing confidential information - digitimeslivemint.comOpenAI fires 3 AI safety researchers for allegedly sharing confidential company informationwsj.comThree OpenAI Researchers Fired for Allegedly Sharing AI Safety Information - WSJ
Read today's full feed at The Newshound →