Fired researchers accuse OpenAI of ‘chilling’ safety efforts

San Francisco: Three former OpenAI security researchers have accused the company of firing them for prioritizing safety over corporate interests, reigniting the debate around AI safety. Mikita Balesni, Tomek Korbak, and Jasmine Wang released a letter to OpenAI on Thursday, expressing concern over the alleged chilling of the company's open culture.

The trio claims they were terminated for mishandling sensitive information outside company procedures, which OpenAI confirmed in a statement. They argue that their firings may be used to justify ending independent auditors' work, which they contend is crucial for maintaining safety. The researchers emphasize that if they can no longer speak out or collaborate with external parties, the potential for a catastrophic AI-related incident increases.

OpenAI maintains that their actions violate company policies, while the terminated researchers stress their commitment to safety and warn of a potential shift in norms within the company.