OpenAI has dismissed three of its researchers following an investigation that found they violated policies related to the handling of sensitive information, the AI company announced on X last Friday.
The statement from OpenAI comes after the terminated researchers, Jasmine Wang, Tomek Korbak, and Mikita Balesni, published a letter outlining the circumstances of their firing. They claimed that their abrupt dismissal could create uncertainty among remaining employees and undermine the culture that previously allowed researchers to raise safety concerns.
“I believe we were terminated because we prioritized safety over OpenAI’s short-term corporate interests,” Balesni stated in a post on X accompanying the letter.
However, OpenAI asserted that the firings were unrelated to safety reporting or the expression of opinions. The company stated, “Our internal investigation revealed a significant breach of trust that goes beyond what is described in the letter they published, and we stand by our decision not to extend their employment contracts.”
OpenAI emphasized that discussions about safety and research happen daily within the company, often in intense and critical debates. They actively encourage these discussions, viewing them as essential for making sound decisions. “We have not dismissed and do not dismiss any employees for expressing their concerns,” the company added.
While OpenAI did not disclose the specifics of the alleged violations, Wang, Korbak, and Balesni noted in their letter that they were not the source of a recent article published in The Information concerning security concerns surrounding OpenAI’s latest AI model, “Astra.”
This announcement coincides with current and former researchers from major AI firms—OpenAI, Google DeepMind, and Anthropic—warning that companies are not taking sufficient measures to protect against the potential negative consequences of developing self-improving AI systems that may become difficult for humans to control.
