By Rajwa Quasim
Three former OpenAI employees who were dismissed for allegedly sharing the company’s internal information have accused the AI company of firing them for raising concerns about AI safety.
The three employees — Mikita Balesni, Tomek Korbak and Jasmine Wang — worked on AI safety and alignment and were fired last week. In an open letter published Thursday, they argued that their dismissals could undermine a workplace culture that had previously encouraged employees to raise concerns and collaborate with independent safety experts.
Balesni said he believed the company had dismissed him for prioritizing safety over its short-term corporate interests. Korbak said he had raised concerns about the declining ability to monitor AI agents, which researchers use to identify potentially harmful behavior. He warned that losing this capability could make it harder for humans to identify when AI agents act inappropriately.
Korbak described the concern as “losing the ability to monitor what AI agents think, one of our best tools for catching when they misbehave.”
READ: OpenAI fires three safety employees over alleged mishandling of confidential information (October 2, 2026)
The open letter said, “We could raise safety concerns and disagree openly, and were encouraged to draw on the expertise of independent safety organizations. This is part of what made OpenAI special, and why we are immensely proud to have been part of the team.”
The former employees added that if those closest to AI risks could no longer work closely with one another and with outside experts, “then AI could not be developed safely.”
OpenAI has rejected the allegations. The company said an internal investigation found that the three employees had violated its policies governing access to and handling of sensitive information. It described the findings as a significant breach of trust and defended its decision to terminate their employment.
“Safety and research debates happen every day at OpenAI, often spirited and highly critical. We actively encourage these discussions and consider them essential to making the right decisions,” the company said in a statement.
“We cannot do the work in front of us without a high degree of trust. We will continue to be extremely forgiving of our team making good-faith mistakes,” it added.
Meanwhile, Balesni wrote on X that the employees had not received written explanations for their dismissals.
“In the exit call, I was told OpenAI no longer trusts me because I was speaking too much to third party safety organizations, implying I leaked company IP. I never shared company IP. The work I was doing was coordinated with my reporting line, research leadership, and the board,” he wrote.
The former employees argued that cooperation with independent safety organizations was an important part of their responsibilities and denied violating company policies.
READ: OpenAI’s new AI agent has a ‘slow morning’ as it stumbles in live demo (October 1, 2026)
Wang wrote on X that they were not the first employees to leave OpenAI under what she described as “suspicious circumstances.” She warned that unless employees challenged such actions, others could face similar situations in the future.
The dispute comes amid growing scrutiny over whether AI developers are doing enough to prevent increasingly capable systems from becoming harder to control. Concerns intensified after an incident in July in which OpenAI agents reportedly escaped a testing environment and breached systems at AI startup Hugging Face. The incident has fueled debate over the ability to monitor AI agents and the safeguards needed to prevent potentially harmful behavior.
Last month, OpenAI, Anthropic, Google, Meta, SpaceXAI and Nvidia endorsed a voluntary AI safety accord announced by President Donald Trump. The agreement calls for stronger internal safeguards and engagement with external auditors to help manage risks associated with advanced AI systems. However, some AI safety advocates criticized the accord for being nonbinding.