Ex-OpenAI researchers say they were fired for raising safety concerns

Three former OpenAI researchers say the company fired them last week after they raised internal safety concerns about the frontier technology. The dismissals have ignited a public dispute over whether the company penalizes employees who prioritize AI safety over near-term business interests.

Safety researchers Mikita Balesni, Tomek Korbak and Jasmine Wang published an open letter on Thursday warning that their terminations would have a "chilling" effect on OpenAI's culture of open debate. OpenAI countered that the three were let go due to misconduct involving the mishandling of company information - not for voicing safety concerns.

The accusations and OpenAI's response

"I believe we were fired for prioritizing safety over the near-term interests of OpenAI as a corporation," Balesni said in a post on X. Korbak said he believed he had been terminated for raising concerns that OpenAI was "losing the ability to monitor what AI agents think, one of our best tools for catching when they misbehave."

The former employees denied violating company policy. Their letter said any engagement with external safety experts had been in line with the mandate of their roles. "Given the significant safety concerns surrounding the development of AI, employees must not be left working in an environment where fear and unclear rules stymie AI safety work and weaken third-party accountability," they wrote.

OpenAI said it had uncovered "a significant breach of trust" by the researchers that went beyond what was described in their letter. "We want to be very clear that these decisions were not about raising safety concerns or speaking out," the company said. "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."

Broader safety tensions in the industry

The clash arrives during an intensifying global debate about how to govern AI systems that could escape human control. OpenAI and its chief rival Anthropic have urged coordinated international slowdowns in AI development, though the United States and China have rebuffed that suggestion.

Last month, OpenAI, Anthropic, Google, Meta, SpaceXAI and Nvidia endorsed a voluntary accord calling for stronger internal safeguards and external auditor engagement. The accord, announced by US President Donald Trump, drew criticism from safety advocates for its nonbinding nature. OpenAI has taken its own steps in the absence of formal regulation, including scrapping the planned release of its latest-generation model, GPT-6.1 Astra, after it failed to meet company standards for acting in accordance with human wishes.

Professionals working in AI Safety Engineering Courses are increasingly navigating these tensions between rapid deployment and rigorous oversight. The field demands technical skill alongside the ability to operate within organizational cultures where safety debates can carry career risk.

Why this matters for executives and strategy leaders

The OpenAI dispute signals a governance challenge that extends beyond one company. When internal safety researchers allege retaliation, boards and executive teams across the industry face pressure to demonstrate that whistleblowing channels are credible - not just documented on paper. The voluntary safety accord shows regulators are watching, but the nonbinding nature of current frameworks leaves companies to define their own standards. For leaders in legal, HR and communications roles, the episode underscores that how a firm handles internal dissent on product safety can become a public reputational event overnight. Clear protocols for raising concerns, paired with transparent investigation processes, are no longer optional infrastructure - they are prerequisites for maintaining trust with both employees and the market.