OpenAI CEO Sam Altman stated in a recent interview that humanity faces a non-zero risk of existential catastrophe from artificial intelligence development before the end of the decade, echoing growing internal and competitive debates over safety standards in the tech sector. Speaking with Fortune editor-in-chief Alyson Shontell, Altman addressed prior industry warnings regarding advanced large language models while competitors like Anthropic and Google weigh proposals to slow down development.
Evaluating Existential Risk Estimates in Artificial Intelligence
When asked if a ten percent probability of human extinction by the end of the decade was accurate, Altman told Fortune that he does not believe that specific figure is correct. However, according to Altman, the exact percentage matters less than the fundamental duty to mitigate harm. “Whether it is ten or eight or six—we have an enormous responsibility and cannot let egos or the pursuit of profit or anything else get in our way,” Altman stated in the interview, adding that developers must act to prevent such outcomes.
The discussion follows public warnings from industry researchers. A former OpenAI researcher who moved to Anthropic recently warned that both companies are playing with human lives by advancing models they privately believe could wipe out humanity by the end of the decade. Anthropic CEO Dario Amodei subsequently published a blog post urging the broader artificial intelligence industry to slow down the training pace of its most powerful models so risk prevention measures can keep pace. Amodei also stated an intention to grant independent auditors access to Anthropic’s facilities.
Corporate Safety Measures and Delayed Public Offerings
OpenAI recently postponed a planned initial public offering. Altman told Fortune that launching a public stock offering right now would be unwise given everything happening regarding safety. Financial analysts, however, noted that the IPO delay was widely anticipated for weeks due to the company’s financial performance rather than solely safety considerations.
Despite differing market motives, major industry leaders expressed support for a deliberate slowdown. Google and OpenAI both welcomed the safety-focused proposals from Anthropic. Elon Musk, founder of xAI, also agreed with the sentiment of caution despite frequently criticizing government-mandated safety regulations.
Contrasting Regulatory Approaches in the United States and Europe
Regulatory philosophy remains a core point of divergence among major artificial intelligence developers operating globally. Musk criticized regulatory frameworks during industry discussions, pointing specifically to the European Union. According to Musk, the regulatory burden in the EU is exceptionally high, creating an environment where many technologies are restricted by default.
In contrast, the United States offers a more permissive regulatory landscape. Home to the world’s leading artificial intelligence firms, the U.S. sector anticipates continued deregulation under the administration of Donald Trump. Whether major artificial intelligence laboratories will successfully coordinate voluntary slowdowns or if market pressures will override safety warnings remains uncertain as firms navigate upcoming product cycles.
>Related reading