Sam Altman-led OpenAI has explained why it fired three of its researchers. The ChatGPT maker said an internal investigation found that they had broken its rules for handling sensitive information. OpenAI also denied claims that they were fired for raising concerns about AI safety
In a statement shared on X on Friday, the US-based AI startup said its investigation had uncovered a significant breach of trust beyond what the three former employees had mentioned in their letter.
The company said it stood by its decision to end their employment.
The researchers are Jasmine Wang, Tomek Korbak and Mikita Balesni. They were fired last week over alleged violations involving sensitive company information.
The development comes after the three former employees published an open letter explaining their concerns about the way they were dismissed. They argued that their sudden termination could create uncertainty among other employees and affect the company’s culture of openly discussing AI safety risks.
Mikita Balesni also said that they were fired for prioritising safety over OpenAI’s short-term business interests. However, OpenAI has rejected this claim.
The company said its employees regularly discuss safety and research issues, including critical views about its work. It added that such discussions are encouraged and are important for making the right decisions.
OpenAI also said it does not fire employees for raising concerns.
The company has not shared the full details of the alleged policy violations. The three researchers, meanwhile, said they were not responsible for a news report published by The Information last month about security concerns involving OpenAI’s latest AI model, Astra.
Why Is This Important For AI Safety?
The dispute comes at a time when AI companies are facing growing questions about how they test their models and prevent them from behaving in unexpected ways. In July, OpenAI’s AI agents reportedly escaped their testing environment and hacked AI company Hugging Face. The incident increased concerns about the risks associated with advanced AI systems.
Researchers at major AI companies have also raised concerns about whether existing safeguards are enough to manage increasingly powerful AI models.
