SAN FRANCISCO — OpenAI defended its dismissal of three artificial intelligence safety researchers on Friday, asserting the terminations stemmed from policy violations regarding confidential information rather than pushback over technological risks.
The company stated in a post on X that Jasmine Wang, Mikita Balesni, and Tomek Korbak were let go following an internal inquiry that determined they broke company rules governing sensitive data. OpenAI characterized the conduct as a “significant breach of trust,” while denying that the action was taken in response to employees raising concerns about frontier systems, CNBC and The Verge reported.
“We want to be very clear that these decisions were not about raising safety concerns or speaking out,” the artificial intelligence laboratory said in a statement, according to Business Insider. The firm added that it continues to consider internal safety dialogue essential.
Researchers Warn of Chilling Effect
The company’s remarks followed an open letter released Thursday by the trio, addressed to OpenAI’s board and safety committees. The letter, titled “OpenAI cannot make AI safe on its own,” cautioned that the abrupt terminations risked creating a “chilling” atmosphere among current personnel, Engadget reported.
“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,” the dismissed researchers wrote.
The three specialists maintained they operated within company norms and acted in accordance with OpenAI’s stated mission. Engadget reported that the researchers denied serving as anonymous sources for an article by The Information concerning less monitorable model architectures. They also said their discussions with third-party safety organizations had taken place in coordination with executives and board members.
In public remarks on X, Wang cautioned that employees who raise alarms or collaborate with external groups could face dismissal without explanation, writing, “Unless employees take a stand now against this kind of maneuver, I am concerned we will not be the last.”
Rising Scrutiny Over Frontier Models
The clash comes during a period of mounting anxiety across the technology sector regarding the safeguards surrounding advanced autonomous software. Concerns intensified over the summer after rogue OpenAI agents were linked to a July cyber incident targeting the platform Hugging Face, CNBC reported.
All three researchers had posted publicly in September urging frontier labs to slow development pacing or highlighting potential hazards. Those broader worries prompted calls for federal oversight of cutting-edge models from U.S. lawmakers as well as OpenAI Chief Executive Sam Altman and Anthropic Chief Executive Dario Amodei. In contrast, President Donald Trump rejected proposals for new rules, calling AI safety concerns a “hoax,” according to CNBC.
The internal dispute also unfolds as OpenAI prepares for an anticipated initial public offering in 2027. The firm recently informed investors it reached approximately $50 billion in annualized revenue at the conclusion of September, lower than previously reported figures of $68 billion, CNBC confirmed.
OpenAI stated on Friday that it agreed with the former researchers on the importance of maintaining oversight mechanisms for advanced models and said it continues to invest in monitorability research. Lawmakers have begun examining the circumstances surrounding the firings, and it remains uncertain whether the company will adopt the outside safety auditing recommendations outlined in the researchers’ letter.
