The AI Industry's Self-Regulation Debate: A Strategy to Avoid Government Control

DeepMind CEO Demis Hassabis recently proposed a US-led international AI standards body modeled after FINRA, the financial industry regulatory authority. The idea is a self-regulatory organization (SRO) that would be federally overseen but industry-funded and run by independent technical experts. Frontier labs would submit their models 30 days before release, with benchmarks updated quarterly, and the body could coordinate a slowdown in development if catastrophic risks emerged. The proposal has drawn broad support from Elon Musk, Sam Altman, Jack Clark at Anthropic, Sundar Pichai, and Satya Nadella.

Why an SRO Instead of Government Regulation

The core argument for an SRO is that government agencies lack the expertise to evaluate rapidly evolving AI models. David Sacks, who discussed the proposal directly with Hassabis, explained that creating a new federal agency—what Anthropic CEO Dario Amodei has called "an FAA for AI"—would be disastrous. The FAA requires type certification for new aircraft designs, a process that takes five to nine years. Even amending an existing certificate takes three to five years. Applied to AI, that timeline would mean models waiting months or years for approval, while China faces no such constraints.