According to reliable sources, on April 12, the artificial intelligence startup OpenAI recently fired two researchers accused of leaking sensitive information, drawing widespread attention in the industry.
Leopold Aschenbrenner and Pavel Izmailov, the two researchers in question, were reportedly integral members of OpenAI's efforts to ensure the societal safety of artificial intelligence. Aschenbrenner, a close ally of OpenAI's Chief Scientist Ilya Sutskever, was involved in AI safety initiatives. Izmailov, on the other hand, specialized in reasoning research and also worked within the security team.
The specifics of what information these two employees leaked remain unclear. However, their dismissal marks the first publicized personnel change at OpenAI since Sam Altman's return to the board in March, following his unsuccessful ousting last fall. Altman's removal had been advocated by Sutskever, among others.
OpenAI, recently valued at a staggering $86 billion through employee stock sales, is engaged in fierce competition with industry giants like Google and Anthropic, aiming to develop the next generation of foundational models that power cutting-edge products like ChatGPT.
Notably, Aschenbrenner was part of OpenAI's Superalignment team, led by Sutskever, dedicated to developing technology to control and manipulate advanced AI, often referred to as "superintelligence," to tackle major human challenges such as fusion power and interstellar colonization.
Before the dismissals of Aschenbrenner and Izmailov, there were evident disagreements among OpenAI staff regarding the safe development of AI. Aschenbrenner was closely associated with the Effective Altruism movement, emphasizing prioritizing the mitigation of potential risks from AI over short-term gains or productivity enhancements.
Sutskever, a key figure in OpenAI's major technological breakthroughs, served on the board during Altman's dismissal and criticized Altman's lack of transparency. However, upon Altman's reinstatement, Sutskever resigned from the board and has since maintained a low profile.
Aschenbrenner, a prodigious researcher who graduated from Columbia University at 19, previously worked at the Future Fund, founded by cryptocurrency exchange luminary Sam Bankman-Fried, focusing on funding innovative projects that "improve humanity's long-term prospects." Aschenbrenner joined OpenAI a year ago.
Requests for comment on the dismissal event and related matters from representatives of Aschenbrenner, Izmailov, and Sutskever, including their attorney Alex Weingarten, have gone unanswered.
Furthermore, it's noteworthy that some former board members involved in Altman's dismissal have ties to the Effective Altruism movement. For instance, Tasha McCauley sits on the board of Effective Ventures, the parent company of the Center for Effective Altruism. Helen Toner held significant roles in open philanthropy projects focused on Effective Altruism. Both departed from OpenAI's board after Altman resumed his CEO position at the end of last year.
