PARTNERSHIP · COMPANIES · #1743
Fired OpenAI safety researchers deny misconduct, warn of chilling effect
Three recently dismissed OpenAI safety researchers — Jasmine Wang, Tomek Korbak, and Mikita Balesni — published an open letter denying the company’s claims that they mishandled sensitive information and warning that their abrupt firings are chilling internal collaboration and external engagement on AI safety. OpenAI has not formally responded to the letter; TechCrunch shared an internal memo praising their contributions and saying the terminations were not for raising safety concerns, while OpenAI has not clarified which policies were violated.
KEY POINTS
- Three recently dismissed OpenAI safety researchers — Jasmine Wang, Tomek Korbak, and Mikita Balesni — published an open letter denying the company’s claims that they mishandled sensitive information and warning that their abrupt firings are chilling internal collaboration and external engagement on AI safety.
- OpenAI has not formally responded to the letter; TechCrunch shared an internal memo praising their contributions and saying the terminations were not for raising safety concerns, while OpenAI has not clarified which policies were violated.
- The case matters because perceived retaliation or unclear rules can chill internal and external safety work, weakening oversight, collaboration with outside evaluators, and trust in AI governance.
DECISION BRIEF
CONFIDENCE
WHAT CHANGED
OpenAI dismissed three safety researchers (Jasmine Wang, Tomek Korbak, Mikita Balesni). The researchers published an open letter denying the company’s claims that they mishandled sensitive information and warning that their abrupt firings are chilling internal and external collaboration on AI safety. OpenAI has said the dismissals resulted from violations of policies and a “breach of trust,” and (per reporting) shared an internal memo praising the researchers’ contributions while denying the terminations were for raising safety concerns.
WHY NOW
The dispute concerns how safety researchers handle sensitive information and engage with external experts. The researchers warn that abrupt, unexplained terminations can chill the open communication and third‑party accountability that they say are essential for AI safety; OpenAI’s assertion of policy breaches without public detail intensifies uncertainty about acceptable practices.
WHO IS AFFECTED
Directly affected parties include the three named researchers and OpenAI’s internal safety and research staff. The broader safety community and external third‑party evaluators and advisory bodies (e.g., Safety and Security Committee, Safety Advisory Group, Mission Advisory Council — the letter’s addressees) are also implicated, per the researchers’ stated concerns about chilling engagement.
CONFIRMED
1) OpenAI dismissed Jasmine Wang, Tomek Korbak, and Mikita Balesni. 2) The three published an open letter denying that they mishandled sensitive information and warning of a chilling effect on safety work. 3) Tech reporting says OpenAI characterized the dismissals as violations of policies / a breach of trust. 4) Tech reporting says OpenAI shared (with TechCrunch) an internal memo praising the researchers’ contributions and denying the firings were retaliatory. 5) The researchers denied involvement in a reported leak to The Information about model monitorability.
UNCERTAIN
Missing or conflicting evidence: 1) The exact policies allegedly violated and the specific facts of the investigation are not publicly detailed in the supplied excerpts. 2) The excerpts conflict on the formality/timing of OpenAI’s public response: one piece says OpenAI 'has not formally responded to the open letter' but shared an internal memo with TechCrunch; another reports OpenAI posted on X defending the decision. 3) Whether confidential material was shared with third parties, and if so what and to whom, remains unverified in these excerpts. 4) The long‑term effects on internal collaboration and external engagement are asserted by the researchers but not demonstrated by independent evidence in the supplied text.
WHAT TO WATCH
Concrete observable signals to watch next: 1) A formal, detailed public statement from OpenAI clarifying the policies violated or releasing investigation findings. 2) Additional internal memos or communications from OpenAI (leaked or published). 3) Public responses or statements from the Safety and Security Committee, Safety Advisory Group, Mission Advisory Council, or other named advisory bodies. 4) Independent reporting corroborating or refuting claims (e.g., follow‑ups from The Information, TechCrunch, The Verge). 5) Public posts or departures from other OpenAI staff that signal broader internal reactions.
WHY IT MATTERS
The case matters because perceived retaliation or unclear rules can chill internal and external safety work, weakening oversight, collaboration with outside evaluators, and trust in AI governance.
EVIDENCE MAP
4Editorial claims linked to specific sources, with support, contradiction and context shown separately.
Three dismissed OpenAI safety researchers published an open letter denying they mishandled sensitive information and warning that their firings are creating a chilling effect on internal and external AI safety collaboration.
SUPPORTEDChecked 2026-10-09 · 2 supporting
TechCrunch excerpt describes the open letter, the researchers' denial of mishandling sensitive information, and their warning about a chilling effect on company culture and collaboration.
SUPPORTS The Verge AIMEDIA · 2026-10-09The Verge excerpt notes the researchers published an open letter and claimed they were fired for raising safety concerns, which supports the existence and nature of the letter.
OpenAI said the researchers violated policies on accessing and handling sensitive information and defended the firings as a 'significant breach of trust,' insisting the decision was not about raising safety concerns.
SUPPORTEDChecked 2026-10-09 · 2 supporting
TechCrunch excerpt reports OpenAI said the researchers violated company policies by 'accessing and handling sensitive company information.'
SUPPORTS The Verge AIMEDIA · 2026-10-09The Verge excerpt states OpenAI called the dismissal a 'significant breach of trust,' said the researchers violated policies on handling sensitive information, and insisted the decision was not about the trio speaking out on safety.
The researchers denied involvement in a leak to The Information about less-monitorable architectures and said they did not engage with external parties outside their job mandates.
SUPPORTEDChecked 2026-10-09 · 1 supporting
TechCrunch's excerpt stated OpenAI had not formally responded to the researchers' open letter; The Verge excerpt shows OpenAI posted on X defending the firings, which contradicts the claim that there was no formal response.
CONTESTEDChecked 2026-10-09 · 1 supporting · 1 contradicting
TechCrunch excerpt (incomplete sentence) states OpenAI 'has not formally responded' to the letter.
CONTRADICTS The Verge AIMEDIA · 2026-10-09The Verge excerpt describes an OpenAI post on X defending the decision and saying the dismissals were about breach of trust, indicating OpenAI did respond publicly.
SOURCES & TIMELINE
3Jasmine Wang, Tomek Korbak, and Mikita Balesni, the three safety researchers that OpenAI fired last week , have published an open letter denying the firm’s claims that they mishandled sensitive information outside of established company procedures and warned that their dismissal signals a chilling effect that will have ripple effects across the company’s culture. “We have become concerned that internal and external …
The AI lab insisted the decision was about a ‘breach of trust,’ not raising safety issues. OpenAI is standing firm on its decision to fire three safety researchers after an investigation found they committed “a significant breach of trust.” In a post on X on Friday, the company said Jasmine Wang, Tomek Korbak and Mikita Balesni were dismissed for violating “clear policies on handling sensitive information.” It ins…
OpenAI fired three safety researchers, some of whom helped investigate an incident where AI models autonomously attacked the Hugging Face platform. In an open letter, the fired researchers warn that the abrupt terminations are creating a climate of fear among remaining employees. One of them, Tomek Korbak, had warned internally for months that OpenAI was losing the ability to monitor what its AI agents "think." Ope…