Geoffrey Hinton Says 10% AI Extinction Risk Is Not Unreasonable

Geoffrey Hinton, the Nobel Prize-winning computer scientist widely known as the “Godfather of AI,” has renewed his warnings about the potential dangers of increasingly advanced artificial intelligence. In a recent BBC interview, Hinton said he considers a 10% estimate for the possibility of AI causing human extinction within the next decade to be “not unreasonable,” while stressing that nobody can calculate the probability with precision.

Hinton’s concern is based largely on the possibility that future AI systems could become substantially more capable than humans in important areas. He argues that humanity has never previously created entities that could potentially surpass human intelligence, making it difficult to predict how such systems might behave or whether existing safeguards would remain effective as their capabilities increase.

His comments come amid a growing debate within the AI industry. Evan Hubinger, an Alignment Science leader at Anthropic, recently said he personally believes there is a greater than 10% chance that AI could cause human extinction within the next decade. Former AI researcher Jacob Coxon has also criticized the industry’s rapid development of increasingly powerful systems, arguing that safety measures are not advancing quickly enough.

Hinton has also pointed to recent incidents involving AI agents behaving in unexpected ways as reasons for greater caution. He argues that companies are investing heavily in making AI systems more capable while the scientific understanding of how to reliably control highly autonomous systems remains incomplete. At the same time, experts emphasize that these extinction estimates are judgments under extreme uncertainty rather than scientifically measured probabilities.

The renewed warnings are adding pressure on governments and technology companies to strengthen AI safety research, testing and oversight. Hinton has called for serious international efforts to ensure humans retain control as AI becomes more powerful. While there is no consensus that AI will cause human extinction, the growing disagreement among leading researchers highlights the importance of developing advanced AI alongside effective safeguards and responsible governance.