DeepMind CEO predicts AGI in years, urges FINRA-style US regulator

Google DeepMind Chief Executive Demis Hassabis said artificial general intelligence could arrive within the next few years, warning that the economic and societal impact will far exceed any previous technological revolution.

The projection carries significant weight for investors and corporate strategists. AGI refers to systems that match the full range of human cognitive abilities, a development Hassabis framed not as a standard tech milestone but as a civilizational turning point comparable to the discovery of electricity or fire.

For markets, the implications are stark. Hassabis argued the technology could dramatically accelerate drug discovery, clean energy development and the creation of advanced materials, effectively removing resource constraints that currently limit economic growth.

“The magnitude of this technology’s impact will be unprecedented, perhaps 10x of the Industrial Revolution at 10x the speed,” he wrote in a post on X on Tuesday.

However, this velocity introduces severe systemic risks. Hassabis noted that advanced AI models already present cybersecurity threats, while more capable, autonomous systems could significantly amplify biological and nuclear dangers.

A FINRA for frontier AI

To mitigate these risks, Hassabis proposed the US establish a dedicated Frontier AI Standards Body. He suggested modelling the organisation on the Financial Industry Regulatory Authority, creating a federally supervised public-private partnership that brings together government agencies, independent experts, open-source representatives and AI companies.

Under this framework, advanced models would undergo rigorous pre-deployment testing focused on high-risk areas like cybersecurity and biological threats. Developers would initially be encouraged to submit models voluntarily before a transition to mandatory compliance.

The proposal also dictates heavier operational requirements for AI laboratories. Companies would need to adopt stronger cybersecurity measures, publish detailed technical documentation and increase capital allocation toward safety research. Furthermore, they would be required to work directly with regulators to patch vulnerabilities identified after deployment.

Global regulatory race

Hassabis’s intervention aligns with mounting institutional alarm over the pace of AI deployment. Earlier this month, United Nations Secretary-General Antonio Guterres cautioned that AI development is moving faster than regulators can effectively oversee, outstripping the capacity of governments to manage associated risks.

“Nobody in the world knows for sure what is going to happen from here, and even the experts disagree,” Hassabis wrote, advocating for cautious optimism over complacency or alarm.

For the private sector, the immediate takeaway is clear: the narrow window to shape AI governance is closing. Companies must prepare for a rapid transition from voluntary safety pledges to a rigid, FINRA-style compliance regime.