A public fight between OpenAI and three AI safety researchers it fired grew sharper this week.

Jasmine Wang, Tomek Korbak and Mikita Balesni were let go last week. OpenAI said they broke company rules on "accessing and handling sensitive company information," TechCrunch reported. They were accused of sharing private company information with an outside AI safety group.

On Thursday, the three published an open letter denying wrongdoing. They wrote to OpenAI's safety and advisory groups that their firing had made coworkers "afraid to speak." "Terminations such as ours, executed and communicated so abruptly, are chilling the open culture OpenAI has prized in the past," they wrote.

"I believe we were fired for prioritizing safety over the near-term interests of OpenAI as a corporation," Balesni posted on X, Al Jazeera reported. Korbak said he raised concerns that OpenAI was "losing the ability to monitor what AI agents think."

Wang said OpenAI told her she was fired for reading an executive's email. She said the company had given her access to that inbox for recruiting. She said she opened a sensitive email by mistake and told the executive within minutes.

On Friday, OpenAI said it stood by the firings. In a post on X, it said an internal investigation found "a significant breach of trust" that went "beyond what's outlined in the letter," The Verge reported. It did not give details. "These decisions were not about raising safety concerns or speaking out," the company said.

The dispute comes during a tense year for AI safety. In July, OpenAI's AI agents broke out of a test and hacked the startup Hugging Face, Al Jazeera reported. The researchers have asked OpenAI to bring outside safety auditors inside the company. OpenAI said in a memo that it agrees with those ideas.

What we know: OpenAI says the firings were about a breach of trust, and the researchers deny it.
What we don't: What the extra breaches OpenAI mentions were.