FRONTIER Act vests sweeping AI oversight in one new Under Secretary

The FRONTIER Act, introduced in the House on July 23 by Reps. Jay Obernolte (R-CA) and Lori Trahan (D-MA), proposes the first comprehensive federal framework for frontier AI. It would require developers to maintain safety frameworks, file transparency reports on new model releases, and report safety incidents. The bill also creates a licensing regime for Independent Verification Organizations (IVOs) — third-party auditors — gives the Secretary of Commerce emergency authority to suspend or restrict models (even internally), and preempts states from enacting their own frontier AI transparency, auditing, and incident-reporting laws.

What critics find striking is that the entire implementation burden falls on a new position — the Under Secretary of Commerce for AI Security — established by a single line in the Definitions section (§2(21)). That individual alone must set minimum framework requirements, design IVO licensing rules (including conflict-of-interest and assessment standards), define “substantial” and “material” modifications, adjust coverage thresholds, run confidential incident-reporting channels, and collect fees to fund the office. Nearly all regulatory duties must be completed within 180 days of enactment. The Under Secretary also receives unredacted versions of public reports — making the role arguably the most powerful AI regulator in the world, yet created with almost no administrative scaffolding.