Three fired OpenAI researchers say they were let go over AI safety concerns; the company disagrees

Three former OpenAI researchers say they were fired because of their work on artificial intelligence safety. Al Jazeera and the BBC report that Mikita Balesni, Tomek Korbak and Jasmine Wang made the claim on Thursday, about a week after they were dismissed. They published an open letter and posted on social media.

Mr Balesni wrote that he believes they were fired for putting safety ahead of the company’s near-term interests. Mr Korbak said he had been raising concerns that the industry is losing the ability to monitor what AI agents think, which he described as one of the best tools for catching bad behaviour. In their letter they warned that the firings could make colleagues afraid to speak, and they said any contact they had with outside safety experts fitted their roles. They denied breaking company policy.

OpenAI disputes this account. The BBC says a company spokesperson told it on 2 October that an investigation had confirmed the three mishandled sensitive information outside set procedures. On Friday, OpenAI repeated this in a note from research leaders, quoted by both outlets, saying the decisions were not about raising safety concerns or speaking out and that the investigation found a significant breach of trust beyond what the researchers described. Al Jazeera adds that the company said safety debates happen every day at OpenAI and that it encourages them.

The BBC reports that OpenAI said it is finalising contracts with third-party safety assessors and will give details in the coming weeks. It also quotes Ms Wang saying she fears others could be pushed out in the same way.

This is a dispute between named former employees and their former employer. We cannot judge who is right, and neither the details of what information was shared nor any internal documents have been made public in the reports we read.