Google DeepMind CEO Advocates for U.S. AI Standards Body

Demis Hassabis, the CEO of Google DeepMind, has called for the establishment of a U.S. Frontier AI Standards Body tasked with reviewing AI models for national security implications prior to their market release. This proposal is in line with President Trump’s Executive Order 14409, which initiated a voluntary 30-day pre-release review framework for certain AI models. However, Hassabis aims to formalize this process into a mandatory requirement through the proposed regulatory body.

Currently, the U.S. framework for AI governance operates on a voluntary basis, with the Commerce Department's CAISI conducting evaluations based on agreements. Hassabis’s suggestion would transition this framework to a mandatory one, reflecting a significant shift in how AI model releases are regulated, particularly concerning national security priorities. This change could also influence market perceptions regarding the federal review of AI models as mandated by Trump.

As it stands, prediction markets indicate a low probability—7%—that Trump will enforce a federal review of AI model releases by July 31. This figure highlights the uncertainty surrounding the timing and scope of potential regulatory changes and their impact on existing executive orders.

Monitoring any official announcements from the White House or the Commerce Department regarding the AI model review process will be crucial. The establishment of the Frontier AI Standards Body could signal a move towards mandatory reviews, which would likely affect the current market odds. Additionally, the outcome of the prediction market concerning Trump’s federal review deadline will provide further insight into the administration’s regulatory approach to AI.