OpenAI denies dismissing three researchers over AI safety concerns

OpenAI said on October 9 that the dismissals of Jasmine Wang, Tomek Korbak and Mikita Balesni were unrelated to their AI safety concerns. The company cited findings from an internal investigation.

In a post on X, the company attributed its decision to part ways with the researchers to a significant breach of trust. The response followed a letter the former employees published a day earlier, addressed to OpenAI's board and safety committees.

The letter's authors expressed concern that internal and public communications about their dismissals were making former colleagues afraid to discuss problems openly. They described freedom to speak as an important part of the previous working culture.

According to CNBC, all three posted on X in September about AI risks or urged developers to balance progress on frontier models with safety. The dispute comes amid reports of cyber incidents involving autonomous AI systems and calls from researchers across several companies for stronger oversight.