DEEPMIND’S CEO MOVES UP HIS AGI DEADLINE — SAYS 2029 IS NOW A REAL POSSIBILITY

Demis Hassabis, the CEO of Google DeepMind, said at Google I/O in late May that he now considers 2029 a genuine possibility for the arrival of artificial general intelligence, pulling his estimate forward from the 2030 to 2035 range he cited just a year ago. He stopped short of calling it a prediction. He called it a real possibility, which from the person running the most funded AI research organization on earth is not a casual remark.

Hassabis said the shift reflects growing confidence that the industry has found the right technical path. He did not elaborate on what that path is, but the statement lands in a very specific context. DeepMind has spent the last 18 months publishing research across protein structure prediction, mathematics, robotics, and general reasoning at a pace that suggests they believe the hard problems are becoming tractable. The AGI framing also matters strategically. When the CEO of DeepMind tells the world AGI might arrive in three years, that is not just a forecast. It is a signal to regulators, investors, competitors, and governments that the acceleration is real and the timeline to act is shorter than most official bodies have been planning for. The EU AI Act just pushed its high-risk enforcement deadline to December 2027. If Hassabis is right, that gives regulators a window of about two years before the rules they are writing may no longer describe the systems they are trying to govern.

Keywords: Demis Hassabis AGI 2029, Google DeepMind AGI prediction, artificial general intelligence timeline, AGI arrival date