US Mulls AI Model Oversight Agency with Tech and Finance Input

The U.S. administration is exploring the creation of a new, independent regulatory agency dedicated to scrutinizing the safety of advanced artificial intelligence models. This initiative, reportedly shaped with input from Treasury Secretary Janet Yellen, aims to involve both the tech industry and financial sector in the review process.

Sources indicate the proposed body would operate akin to the Financial Industry Regulatory Authority (FINRA), a self-regulatory organization funded by the industry but overseen by the Securities and Exchange Commission (SEC). White House Chief of Staff Susie Wiles is currently reviewing the plan.

This move comes after recent instances where U.S. national security concerns led to temporary halts or significant modifications of cutting-edge AI models. Companies like Anthropic faced access restrictions for their models, while OpenAI had to revise its latest 'Sol' model before its release. These actions have reportedly frustrated tech leaders who argue that government interventions have been excessive compared to the perceived security risks.

The proposed regulatory framework seeks to offer greater certainty to AI developers and address frustrations over inconsistent government oversight. By allowing Silicon Valley and Wall Street firms a more significant role in setting AI safety standards, the administration hopes to balance security imperatives with the rapid pace of AI innovation. The plan aligns with suggestions from figures like Google DeepMind CEO Demis Hassabis, who has advocated for an industry-funded, expert-led review board operating under government supervision.