Preemption Isn’t Governance. We Need a Federal Framework for AI.

In this article for Public Knowledge, author Nat Purser argues that a recent executive order aimed at preempting state-level AI regulation is misguided, as it prioritizes industry concerns about regulatory fragmentation over the need for responsive and localized governance. While acknowledging that a patchwork of state laws can create compliance challenges, the author contends that these concerns do not justify undermining states’ ability to regulate a rapidly evolving and impactful technology.

The article explains that the executive order directs federal agencies to challenge state AI laws through litigation, regulation, and policy pressure, while also attempting to leverage federal funding—such as unused broadband deployment funds—to incentivize state compliance. However, Purser notes that these mechanisms face significant legal and political obstacles, as executive orders alone cannot override state law, and efforts to repurpose congressionally allocated funds or expand agency authority are likely to be contested.

At the congressional level, the push for broad federal preemption has already encountered strong bipartisan resistance. Attempts to impose a sweeping moratorium on state AI laws have failed, reflecting divisions within both political parties and widespread concern from stakeholders, including state governments and advocacy groups. The author emphasizes that any viable federal preemption effort would require meaningful compromise and a clear regulatory framework to gain sufficient legislative support.

Purser continues, arguing that preemption can serve the public interest only when paired with robust federal protections addressing the same issues being preempted. Rather than eliminating state authority without offering alternatives, policymakers should pursue targeted, issue-specific preemption alongside comprehensive federal regulation. This approach would preserve states’ role as regulatory innovators while ensuring national consistency where necessary, fostering both public trust and effective AI governance…

CONTINUE READING…