OpenAI defends firing three safety researchers amid oversight dispute
OpenAI dismissed safety researchers Tomek Korbak, Mikita Balesni and Jasmine Wang in early October and on Oct. 9 defended the decision, saying an internal investigation found they mishandled sensitive information outside company procedures and broke its trust; the company denied the firings were about raising safety concerns or speaking out. In an open letter, the three disputed that account: Korbak said he was told verbally he was fired over how he communicated with the independent evaluator METR, Balesni said he was suspected of leaking company IP because he spoke with third-party safety organizations (which he denies), and Wang said her dismissal involved access to an executive's email. An Understanding AI commentary ties the episode to OpenAI's system card statement that GPT-6 Astra "shows a substantial decrease in chain-of-thought monitorability compared to previous models," and describes Astra as using a recurrent depth architecture that makes outside monitoring of its reasoning harder.
The dispute tests OpenAI's stated commitment to independent safety evaluation, since employees' contacts with outside evaluators are at the center of the case, while the same reporting notes that a new flagship model's chains of thought are less monitorable.
We may earn from vendors via affiliate links or sponsorships. This might affect product placement on our site, but not the content of our reviews. See our Terms of Use for details.
OpenAI says three safety researchers breached its trust. The researchers say their firings could discourage employees from raising concerns about AI safety.
On Oct. 9, OpenAI defended its earlier dismissal of safety researchers Tomek Korbak, Mikita Balesni and Jasmine Wang, saying an internal investigation found that they violated company policies governing sensitive information. The former employees dispute that account, maintaining that their work with outside safety organizations fell within their responsibilities.
The dispute raises questions about how OpenAI balances confidentiality with independent safety evaluations, particularly as businesses increasingly rely on advanced AI systems.
Why OpenAI fired the researchers — and why they disagree
OpenAI told TechCrunch that its investigation uncovered misconduct extending beyond communications with external evaluators. The company denied that the dismissals were retaliation for raising safety concerns but has not publicly detailed the specific policies allegedly violated.
#OpenAI#Tomek Korbak#Mikita Balesni#Jasmine Wang#METR#GPT-6 Astra#Hugging Face#Fable 5.1#Gemini 4 Argon#Google#Anthropic