October 9, 2026, (Inside AI) — OpenAI has defended its decision to terminate three AI safety and alignment researchers last week, citing a "significant breach of trust" that it says went beyond what the former employees disclosed in a public letter. The company denied that the firings were retaliation for raising safety concerns, but the episode has ignited fresh debate over whether the world's most prominent AI lab can police itself on safety matters.
The three researchers, Mikita Balesni, Tomek Korbak, and Jasmine Wang, were fired after an internal investigation. They had been instrumental in probing the Hugging Face hack linked to OpenAI, a security incident that raised questions about the vulnerability of AI supply chains. In a letter addressed to OpenAI's Safety and Security Committee on October 8, they argued that advanced AI cannot be developed safely unless researchers can collaborate freely with independent experts. Hours later, OpenAI pushed back, saying its investigation found policy violations related to handling sensitive information.
The dispute centers on whether the researchers' external communications with third-party safety organizations crossed a line. Korbak said he was told verbally that he was fired because of how he communicated with METR, an AI safety nonprofit that partnered with OpenAI to investigate the Hugging Face incident. "I was told verbally I was fired because of the way I communicated with METR. No details on what I said or did or when. No other reasons were given, and nothing was put in writing. To be clear, talking to METR was my job," Tomek Korbak, former OpenAI safety researcher, wrote on X.
Balesni said OpenAI told him he was speaking too much to third-party safety organizations. He understood the company to be implying that he had leaked intellectual property, an allegation he denied. Wang warned that unless employees take a stand, more firings could follow. "The message to everyone still at OpenAI is clear: raise concerns or work closely with outside safety groups, and you could be next, without being told why. You can't build AGI safely if the people closest to the risks are afraid to speak," Jasmine Wang, former OpenAI researcher, wrote on X.
OpenAI rejected the retaliation narrative. "Our internal investigation uncovered a significant breach of trust beyond what's outlined in the letter they published, and we stand by the decision to not continue their employment," the company said in a statement. It added that the decisions "were not about raising safety concerns or speaking out" and that internal debates about safety are encouraged, including those that are "spirited and highly critical." OpenAI did not specify what the alleged violations were.
The company also said it is finalizing contracts with third-party safety assessors and will announce details in the coming weeks. It reiterated its commitment to working with independent safety organizations. "We are deeply sad about this outcome," OpenAI said, adding that it valued the former employees' contributions to AI safety and their willingness to challenge ideas. "We have always encouraged that and always will."
The firings come amid growing scrutiny of OpenAI's safety practices. In recent months, the company has faced criticism from former employees, regulators, and researchers who argue that commercial pressures are undermining its original safety mission. The departure of key safety personnel has raised questions about whether OpenAI can maintain its commitment to developing AI systems that are safe and beneficial.
The three researchers had worked on alignment, the technical challenge of ensuring AI systems behave as intended. Their work on the Hugging Face hack was seen as critical to understanding how attackers might exploit vulnerabilities in AI models. The incident, which involved unauthorized access to a Hugging Face repository linked to OpenAI, highlighted the risks of open-source AI development and the need for robust security measures.
In their letter, the researchers stressed that monitoring advanced AI models and understanding their behavior is essential for safety. "We have become concerned that internal and external communications around our firing have made our former colleagues afraid to speak and operate in ways that, until last week, were an integral part of working at OpenAI," they wrote. They argued that the freedom to collaborate with independent experts without fear is itself a safety mechanism.
OpenAI's response suggests a fundamental disagreement over what constitutes appropriate external communication. The company has long emphasized its commitment to safety, but critics say its actions tell a different story. The episode echoes earlier controversies, including the 2023 boardroom drama that briefly ousted CEO Sam Altman, which was partly fueled by concerns over the pace of commercialization versus safety.
For now, the three former researchers are speaking out, and OpenAI is standing firm. The company's forthcoming contracts with third-party assessors may offer a chance to rebuild trust. But the firings have already sent a chilling message to some in the AI safety community. As Wang put it, the fear of speaking up could make it harder to build AGI safely. Whether OpenAI can convince the world otherwise remains an open question.