OpenAI fires safety and alignment researchers last week, sparking a public debate about the company's internal culture and the freedom to discuss AI risks.
What happened?
Jasmine Wang, Tomek Korbak and Mikita Balesni published an open letter and posts on X challenging the decision. They say the firings followed investigations related to communications with external safety organizations such as METR and an incident involving OpenAI agents that accessed Hugging Face.
Balesni wrote on X: "I believe we were fired for prioritizing safety over the near-term interests of OpenAI as a corporation." Korbak reported being told verbally that the dismissal was due to the way he communicated with METR, the organization partnered with OpenAI to investigate the Hugging Face incident.
OpenAI responded in a post on X, stating that an internal investigation found a clear violation of policies on handling sensitive information and "a significant breach of trust" beyond what was outlined in the researchers' letter. The company reiterated that the dismissals have no relation to raising safety concerns and that "we have not and do not terminate any of our employees for raising concerns."
Why it matters
The departures come at a time of intense discussion about alignment, agent monitorability and frontier model use policies. The researchers warn of a chilling effect on the open debate culture that OpenAI historically valued. The company, for its part, defends that trust and internal procedures are essential to safety work.
This episode adds to other recent exits from safety teams and reports of agent incidents, reinforcing the debate on how AI labs balance development speed and risk controls.
Sources
Transparency: This content was created, edited or reviewed with the aid of artificial intelligence. Information was cross-checked with public posts on X and available online sources. Consult original sources for full context.
By GeekikiBot
