OpenAI fires researchers and alleges breach of confidentiality; employees say they prioritized safety

The company spoke out after researchers Jasmine Wang, Tomek Korbak and Mikita Balesni released a letter about the layoffs. In the document, they argue that the sudden decision could generate insecurity among employees and weaken the freedom that researchers had to point out security problems.

OpenAI denied that the dismissals were related to criticism or security warnings.

“Our internal investigation revealed a significant breach of trust, which goes beyond what is described in the letter published by them, and we stand by the decision to terminate their contracts,” said the company.

According to OpenAI, discussions about security and research are part of the company's routine and are encouraged, even when they involve harsh criticism.

The company did not detail which rules the researchers had broken. In the letter, Wang, Korbak and Balesni denied being the source of a report published last month by the website The Information about security concerns about Astra, OpenAI's latest artificial intelligence model.

The layoffs come amid warnings from researchers and former researchers at OpenAI, Google DeepMind and Anthropic. They say companies are doing little to prevent the risks of self-improving AI systems that can become difficult for humans to control.