US Senators pitch a new AI ‘super‑regulator’ to rein in Big Tech and frontier models

On September 23, US Senators Michael Bennet and Peter Welch, both Democrats, unveiled a proposal to create a new federal agency dedicated to regulating frontier AI models and the largest tech platforms. Their pitch is that today’s patchwork of watchdogs — such as the FTC and FCC — is too fragmented to cope with the speed and scale of generative AI and social media.

The envisioned commission would be able to designate “systemically important” AI models and digital platforms, subjecting them to stronger reporting requirements and safety obligations. The idea borrows from post‑financial‑crisis bank regulation: if a model or platform poses outsized risks to competition, privacy, or civic life, regulators should have tools to demand changes and impose penalties.

Skeptics worry a new agency could slow innovation or simply add bureaucracy, especially given the Trump administration’s reluctance to embrace stricter AI rules. Turning the proposal into law would require bipartisan support and intense negotiations with industry. Even so, the fact that sitting senators are now arguing for an AI‑specific super‑regulator marks a new phase in the global debate over how far governments should go in policing advanced AI systems.