Google DeepMind’s Demis Hassabis Warns AGI Is ‘A Few Short Years Away,’ Proposes FINRA-Style Regulatory Body

The race toward artificial general intelligence (AGI) is no longer a distant science-fiction horizon; it is an imminent reality that could fundamentally restructure human civilization.

In a stark warning and call to action published Tuesday on X, Demis Hassabis, the Google DeepMind CEO, declared that AGI is “probably only a few short years away.” Far from being a mere incremental upgrade to our digital lives, Hassabis argues that the arrival of human-level machine intelligence will be a civilizational pivot point on par with humanity’s most elemental discoveries.

“When we look back on this time in the decades to come, I think we will realise we were standing in the foothills of the singularity–nothing less than the dawning of a new age for humanity,” Hassabis wrote.

The term AGI refers to the point when computers can understand, learn, and perform a wide range of tasks as well as or better than humans. While previous technological leaps like the internet or mobile computing revolutionized how we access and share information, Hassabis believes AGI belongs in an entirely different category of human achievement.

“It is much more akin to the discovery of electricity or fire,” he noted. “If you stop to think about it, we’ve essentially found a way to make sand think. It’s miraculous.”

Yet, this miraculous leap brings unprecedented existential and structural risks. Hassabis warned that the pace of AI development is currently outstripping society’s capacity to comprehend, let alone govern, the resulting hazards. Today’s frontier models already present tangible cybersecurity threats, but Hassabis cautioned that future iterations could introduce catastrophic biological, nuclear, and national security risks.

As these models become increasingly agentic—capable of executing complex, multi-step workflows independently—and begin to undergo recursive self-improvement, the question of human control becomes paramount.

“On the horizon, we will need robust safeguards to maintain control of increasingly agentic, recursively self-improving systems–and tackle unknown issues that will only become clearer over time,” Hassabis wrote.

The Convergence of AI, Agentic Systems, and Web3

The transition to highly autonomous, agentic AI systems is already reshaping the broader technology landscape, including the digital asset ecosystem. In the blockchain space, developers are increasingly building decentralized infrastructure to support autonomous agents. Because these AI agents lack legal identities or traditional bank accounts, decentralized protocols and smart contracts are emerging as the native financial and coordination rails for machine-to-machine transactions.

However, the centralization of the underlying frontier models remains a point of intense debate. While centralized tech giants build massive data centers to train proprietary systems, decentralized AI networks are attempting to democratize access to compute and model training. The fear among many decentralized open-source advocates is that heavy-handed regulatory frameworks could inadvertently centralize control of AGI in the hands of a few corporate gatekeepers.

A Self-Regulatory Proposal: The FINRA Model

To mitigate these risks without stifling innovation, Hassabis proposed the creation of a U.S. Frontier AI Standards Body. Rather than relying on a slow-moving federal bureaucracy, Hassabis suggested modeling this new entity after the Financial Industry Regulatory Authority (FINRA), a private, self-regulatory organization that oversees U.S. brokerage firms.

Under this proposal, the new standards body would operate as a federally supervised public-private partnership. It would be funded primarily by the AI industry itself and staffed by independent technical experts alongside open-source representatives. This body would be tasked with dynamically evaluating the capabilities and safety profiles of frontier AI models before they are deployed to the public.

“The rapid progress we’re seeing in AI requires a new approach to testing frontier AI model capabilities that is dynamic, adaptable, and rigorous,” Hassabis explained. “The US is well positioned, given its economic and technical standing, to take the first step in developing such a framework.”

A Chorus of Warnings and Divergent Paths

Hassabis is not alone in his urgency. The broader AI sector has spent much of the period following the late 2022 public launch of ChatGPT warning that the timeline to AGI is collapsing.

In January 2026, Anthropic CEO Dario Amodei warned that human-level AI could emerge within one to five years, asserting that global governments are severely underestimating the velocity of this technological shift. Amodei has argued that the industry needs robust safety standards akin to those enforced by the Federal Aviation Administration (FAA) to manage the risks of highly capable models.

Hassabis himself previously sounded the alarm in June, predicting that AGI would arrive by 2030 and warning that society has “not long to prepare.”

This collective anxiety has reached the highest levels of government. In May 2023, during a landmark hearing before the U.S. Senate Committee on the Judiciary, OpenAI CEO Sam Altman advocated for the creation of a federal agency tasked with licensing powerful AI systems and mandating independent safety audits.

More recently, last month, President Donald Trump signed an executive order establishing a voluntary framework designed to review advanced AI models prior to their public release.

The Window of Opportunity

As the debate over AI governance intensifies, Hassabis emphasized that the global community has a narrow, high-stakes window to establish common safety standards before autonomous intelligence surpasses human capabilities.

“The future is not yet written, we must use this precious window before AGI arrives to shape this technology for the benefit of all humanity,” Hassabis urged. “What we collectively do now will determine how the next phase of civilisation unfolds. By safely stewarding AGI into the world, we can enter a new golden age of scientific discovery and progress, and usher in a bright future of incredible human flourishing.”