Science & Tech · AI · 2 days ago
Three former OpenAI safety staff dispute reasons for their firing
OpenAI fired three safety-team employees, including two who worked on an investigation involving Hugging Face.
The former staff deny wrongdoing and say they were fired for their communications with an outside auditor or for an email-access issue.
OpenAI says it found policy violations, but has not given specific reasons for the dismissals.
The dispute matters because the former employees say independent safety checks and open discussion about risks should continue.
OpenAI dismissed three safety-team employees: researchers Tomek Korbak and Mikita Balesni, and program manager Jasmine Wang.
Korbak and Balesni had worked on OpenAI’s investigation of the Hugging Face incident, and Korbak was the technical contact for METR, the firm asked to audit the incident.
In a letter published on Oct. 8, the former employees denied leaking concerns about the monitoring of OpenAI’s Astra model and denied wrongdoing in their communications with METR.
OpenAI said its investigation found a pattern of misconduct and policy violations involving sensitive information, but it has not publicly specified the reasons for each dismissal.
The former employees said the dismissals could discourage staff from raising safety concerns or working with outside auditors, while OpenAI stood by its decision.
- Who
- OpenAI and former safety-team employees Tomek Korbak, Mikita Balesni and Jasmine Wang
- What
- OpenAI dismissed three employees, who disputed aspects of the company’s account and raised concerns about safety oversight and workplace openness.
- When
- The Wall Street Journal reported the dismissals on Oct. 1; the former employees published their letter on Oct. 8. The article was published Oct. 9, 2026.
- Where
- OpenAI; the article does not state a specific location.
- Why
- OpenAI cited alleged misconduct and policy violations involving sensitive information; the former employees gave differing explanations and disputed wrongdoing.
OpenAI
Former employees
Reason for the dismissals
OpenAI
OpenAI said an internal investigation found a pattern of misconduct and violations of its policies for handling sensitive information, and said it stood by the decision to end their employment.
Former employees
The employees challenged the company’s account. Korbak said he was told the way he communicated with METR was the reason for his dismissal; Wang said her dismissal concerned access to an executive’s email that she said had been granted for recruiting; Balesni said they were fired for prioritizing safety.
Communications with METR and outside auditors
OpenAI
OpenAI did not link the dismissals to communications with METR or to the Hugging Face incident.
Former employees
The former employees denied wrongdoing in their communications with METR and urged OpenAI to maintain third-party safety audits and auditors’ access.
Safety culture and speaking up
OpenAI
OpenAI stood by its decision to dismiss the employees and said its investigation uncovered a significant breach of trust.
Former employees
The former employees said the dismissals and the way they were communicated could chill open discussion and make employees afraid to raise concerns.
individuals mishandled sensitive information outside established company procedures, violating our policies and breaking the trust essential to our work.
Terminations such as ours, executed and communicated so abruptly, are chilling the open culture OpenAI has prized in the past.
Talking to METR was my job.
Our internal investigation uncovered a significant breach of trust beyond what’s outlined in the letter they published and we stand by the decision to not continue their employment
The Information published an article about concerns over monitoring OpenAI’s Astra model, which the former employees later denied leaking.
The Wall Street Journal first reported the dismissals.
The three former employees published a joint letter responding to accounts of their dismissals.
Fortune reported that the former employees and OpenAI had issued further statements about the dispute.
- Employees dismissed
- Three: Tomek Korbak, Mikita Balesni and Jasmine Wang
- Safety auditor
- METR, which OpenAI asked to conduct an independent audit of the Hugging Face incident
- Former employees’ letter
- A four-page letter published Oct. 8
- Initial report of dismissals
- The Wall Street Journal reported them on Oct. 1
- Model cited in the dispute
- Astra
Sources
Related news
Microsoft president calls for global AI safety rules
UK to strengthen cyber defences and press AI firms to improve safety
Grok offers limited free access, while paid plans add higher limits and tools
Nadella says advanced AI needs human-controlled emergency brake
Anthropic AI submits false murder tip to Philadelphia police
Anthropic bars repeated cruelty toward its AI models
OpenAI fires three safety researchers after dispute over sensitive information
Nikon disqualifies AI-assisted microscope video and names a new winner
Tata Communications shifts from connectivity to digital infrastructure and AI readiness
Bill Gates warns AI could drive more cyberattacks in the coming year
Anthropic cuts internet access for all internal AI evaluations
