Ex DeepMind engineer joins chorus warning AI could wipe out humanity

A former Google DeepMind engineer became the latest artificial intelligence researcher to warn publicly that the technology could threaten humanity’s survival, adding his voice on Monday to a widening debate over whether the industry is moving too fast to manage the risks it creates.

Bilal Chughtai, who worked as a research engineer on artificial general intelligence safety and alignment at DeepMind before leaving the company in July, posted his warning on social media platform X. He wrote that he genuinely believes AI has the potential to kill everyone, and that the window to prevent that outcome may be closing. Chughtai co-authored several research papers during his time at DeepMind, according to his LinkedIn profile. Google did not respond to a request for comment outside normal business hours.

Chughtai said he remains hopeful that AI can still be developed safely, but argued that doing so will require the industry to coordinate rather than continue what he described as a frantic race between competing companies. He said society needs AI development to proceed at a pace it can actually manage, one that allows emerging risks to be identified and addressed before they cause serious harm. His comments follow the resignation last week of Jacob Coxon, a researcher at Anthropic, who said he was stepping down in part because people building AI systems genuinely believe the technology could kill everyone by the end of the decade. Another Anthropic researcher, Evan Hubinger, responded to Coxon’s remarks by saying he agreed and that he estimates a greater than 10 percent chance that AI could cause human extinction within the next ten years.

The wave of warnings has pushed the debate over AI risk further into public view. Anthropic chief executive Dario Amodei called over the weekend for the pace of advanced AI development to slow down, a position that drew unusual support from figures who do not typically align on technology policy, including Elon Musk of SpaceX and xAI and OpenAI chief executive Sam Altman. Their joint concern has sparked a broader argument within the industry over what critics call AI doomerism, the view that catastrophic risk from advanced systems deserves far more weight than it currently receives in corporate and government decision making.

Not everyone in Washington has embraced the warnings. President Donald Trump has dismissed the concerns raised by AI industry leaders, describing the push for regulation on social media as a hoax. His comments stand in contrast to a growing number of researchers inside the field’s most prominent labs who say they now assign meaningful probability to catastrophic outcomes from the technology they helped build, a shift from a decade ago when most scientists in the field treated existential risk from AI as a distant and largely theoretical concern.

Chughtai’s specific worry centers on artificial general intelligence, systems that could eventually match or exceed human ability across the full range of cognitive tasks. Alignment research, the field he worked in, focuses on ensuring that such systems pursue goals consistent with human values and interests, though no proven method yet exists for guaranteeing that outcome once systems become sufficiently advanced. That uncertainty is part of what has driven a string of departures and public statements from safety focused researchers in recent months, as concern mounts that commercial competition among AI developers is outpacing the industry’s ability to manage the risks those same companies acknowledge.