Future of Life Institute publishes the 'Statement on Superintelligence'

The Future of Life Institute published the “Statement on Superintelligence,” a one-sentence pledge calling for “a prohibition on the development of superintelligence” that would not be lifted until there was broad scientific consensus that it could be done safely and controllably, and strong public buy-in. More than 700 people signed initially, a list the organisers said spanned AI researchers, faith leaders, policymakers and artists.

The signatory list was notable for crossing lines that AI-safety statements had not previously crossed. Alongside AI researchers Yoshua Bengio and Geoffrey Hinton, Apple co-founder Steve Wozniak, Virgin founder Richard Branson and several Nobel laureates, it included right-wing media figures Steve Bannon and Glenn Beck, former US national security adviser Susan Rice, and public figures including Prince Harry and Meghan Markle. Organisers and press coverage presented the breadth as evidence that concern about racing to build superintelligence without safeguards was no longer confined to one political or professional camp. Sam Altman and Microsoft AI’s Mustafa Suleyman were among the notable figures who did not sign; commentary quoted Suleyman’s own prior public position — that AI “shouldn’t be inventing” superintelligence “until we can prove unequivocally that it is” safe — without his name appearing among the signatories.

The statement was widely described as a successor to the 2023 “Pause Giant AI Experiments” letter, which the same organisation had circulated after GPT-4’s release and which had called for a six-month moratorium on training runs beyond GPT-4’s scale. That earlier pause was not observed by any major lab. The 2025 statement made no comparable request for an immediate halt, instead asking for a conditional prohibition contingent on future proof of safety — a framing organisers said reflected two and a half years of unslowed frontier development since the first letter went unheeded.