DeepMind CEO Warns AGI Could Arrive by 2029, Calls for Urgent Government Action
Published on: May 27, 2026
DeepMind CEO Demis Hassabis recently conveyed a heightened sense of urgency about the prospect of artificial general intelligence. Speaking at Google’s I/O developer conference and then to the press, Hassabis stated that humanity may currently be standing in the “foothills of the singularity,” and that AGI could emerge as early as 2029, sooner than his previous projections targeting around 2030.
Hassabis emphasized the growing confidence within the AI research community that the technical path toward AGI is becoming clearer. He suggested that recent advances in agentic AI—systems capable of self-direction and recursive improvement—have accelerated this timeline, prompting him to lower his AGI estimate by a year.
Beyond prediction, Hassabis underscored the need for proactive governance. He praised tentative steps by the U.S. federal government toward prioritizing AI safety, including proposals for pre-release testing of new models, but he warned that public and economic institutions must prepare more broadly for the societal transformations AGI could bring.
Hassabis voiced concern that discussions about AGI’s broader societal impacts remain confined within tech circles. He called for expanded engagement from governments, economists, and the general public to address not only technical safety but also economic, legal, and ethical implications as AI systems become more autonomous and capable.
One key milestone Hassabis highlighted is the onset of recursive self‑improvement, where AI systems accelerate their own development. He noted that such capabilities could dramatically shorten development cycles, increasing the risk that society will be unprepared if governance and safety measures lag behind technical advances.
Hassabis’s remarks have catalyzed renewed attention to AGI timelines and preparedness. By naming 2029 as a plausible horizon—and calling AGI arrival a matter of urgency rather than speculation—he has reframed the public conversation about how much time remains and intensified calls for global coordination on AI safety and regulation.
No comments yet.