Sam Altman has warned that artificial intelligence could go “very badly” in two distinct ways, as the OpenAI chief called for the technology’s rapid progress to be slowed.
Writing on X, Mr Altman said the first danger was that humanity could “lose control of the future of AI”, which he described as “unacceptable”.
“We are unapologetically on Team Humanity and AI must always serve people,” he said, arguing that safety and alignment measures must remain ahead of advances in AI capabilities.
His second concern was excessive concentration of power. Mr Altman warned that if a highly powerful AI system were used by one individual or company to impose its worldview on everyone else, “the results could be extremely dystopian”.
He said he was working with other executives to moderate the pace of development, while stressing that this did not mean bringing progress to a halt.
“When we talk about ‘pacing,’ we don’t mean ‘stopping’,” Mr Altman said in an earlier post. “Progress has been rapid and will continue to be. But it should be slower than it otherwise could be.”
He said the cost of slowing down would be worthwhile, adding that American competition could not justify recklessness or allow AI capabilities to outstrip safety and monitoring.
“The world deserves confidence that American companies developing increasingly capable AI will act responsibly,” he said. “Every frontier lab must deliver on this.”
Whistleblower warns of ‘self-improving superintelligence’
Mr Altman’s comments followed the resignation of Jacob Coxon, a former researcher at both OpenAI and Anthropic, who accused the companies of failing to act responsibly.
Mr Coxon said they were “racing straight to self-improving superintelligence and gambling with our lives”. He urged the public not to underestimate technology that he said could eventually hack anything, transform entire fields overnight and acquire real power and resources.
He also claimed that people developing AI believed it could kill everyone by the end of the decade, insisting that the warning was not a marketing stunt.
Superintelligence refers to a point at which an artificial system becomes more powerful than any individual, company or nation.
Geoffrey Hinton, the Canadian researcher often called the “Godfather of AI”, has similarly warned that superintelligent systems could lead to human extinction.
“We would be very foolish to develop superintelligence now, when there is no scientific consensus it can be developed safely and controllably,” Dr Hinton said. “Losing control over AI smarter than ourselves could be catastrophic and could even lead to human extinction.”
