Google DeepMind CEO Demis Hassabis stated that while he predicts the realization of artificial general intelligence (AGI) will occur around 2030, he is also considering the possibility of it happening as early as 2029.

Google DeepMind CEO Demis Hassabis, who attended Google I/O 2026, Google's annual developer conference, made headlines by stating that the realization of artificial general intelligence (AGI) , capable of thinking like a human, 'could happen as early as 2029.'

Google DeepMind CEO Demis Hassabis says we're close to AGI
https://www.axios.com/2026/05/26/deepmind-ceo-demis-hassabis

AI: Google DeepMind CEO Demis Hassabis steps up on AGI. AI-RTZ #1099
https://michaelparekh.substack.com/p/ai-google-deepmind-ceo-demis-hassabis

In an interview with technology media outlet Axios at Google I/O 2026, Google DeepMind CEO Hassabis revealed his prediction that AGI could be available by 2030, or even as early as 2029. Hassabis explained that this prediction 'reflects a growing confidence that the industry has found the right technological path.'

Furthermore, CEO Hassabis reportedly stated that 'humanity is standing at the foot of the singularity , and society has only a few years left to prepare for AGI.'

In addition, CEO Hassabis stated that 'next-generation AI agents should be seen as a social stress test against far more powerful systems (AI) that will emerge in the future.' As a concrete example, CEO Hassabis cited Anthropic's Mythos, saying, 'The fact that Anthropic's Mythos has the power to catch businesses and governments off guard shows that we are not yet prepared for the rapid evolution of these systems. This was probably a good warning shot.'

Mythos is a security AI developed by Anthropic that infers how to find and exploit software vulnerabilities. Anthropic considers the potential for Mythos to be misused to be a serious problem and therefore only provides it to a limited number of users. In fact, Mythos has discovered numerous potential vulnerabilities .

Anthropic develops 'Claude Mythos Preview,' an AI with extremely high cyberattack capabilities, and has also launched 'Project Glasswing,' which will provide a preview version to Microsoft, Apple, and others - GIGAZINE

Hassabis CEO mentioned that he deliberately chose these words to urge government officials, economists, and the general public to prepare more urgently for increasingly powerful AI. He cited a proposed executive order in the U.S. requiring government testing before new AI is released, saying, 'The U.S. government's provisional step towards prioritizing safety is a step in the right direction.'

Dozens of MAGA supporters demand that President Trump mandate government testing and approval before the release of powerful AI models - GIGAZINE

In addition, CEO Hassabis also stated in an interview with Rowan Chang that he believes AGI will be realized around 2030.

Demis Hassabis on AGI by 2030, Curing Every Disease, Life After AGI, and More - YouTube