Tucked into the reconciliation “megabill” House Republicans passed last month is a provision calling for a 10-year moratorium on state-level AI regulation. The move puts federal preemption at the center of a fierce debate surrounding this evolving technology. While the Progressive Policy Institute has been highly critical of the GOP’s fiscally irresponsible bill, we hope the idea of a moratorium survives the legislative gauntlet.
PPI believes the budget reconciliation process should not be used to enact non-budget policies, such as a moratorium on state AI regulation, and is concerned that a 10-year freeze is too long. However, a shorter, three-year pause — enacted through regular order — would give Congress the opportunity to develop a comprehensive federal framework for regulating AI.
That’s in America’s national interest because with AI growing exponentially, it is essential to get regulation right. In the absence of federal action, the states are racing to enact their own
laws governing the uses of AI — hundreds of bills have been introduced on the topic in the last year alone. That could lead to a confusing mess of partial, duplicative, and conflicting rules, which could hamper AI development in the United States, while competitors for high-tech leadership, such as China, forge ahead.
A moratorium on state AI regulation is crucial because allowing states to choose their own regulatory approaches comes with high costs that harm both consumers and businesses. The recent history of privacy regulation in the United States, which resembles the trajectory of AI regulation today, holds important lessons about these costs. Much like AI today, as Congress stalled, dozens of states passed their own laws regulating privacy, each with its own approach and requirements. As this paper will explore, when Congress later attempted to pass federal privacy bills in 2022
and 2024, these efforts failed because of the proliferation of differing state laws, which made it difficult to reach a political consensus. As a result, citizens and businesses face a web of duplicative and costly state rules that still leave many Americans with no privacy protections.
The country’s experience with privacy regulation holds two important lessons for lawmakers looking to regulate AI. First, state-level regulation leads to balkanization, which directly stifles innovation and increases costs for all parties, all while leaving Americans in states without laws unprotected. Second, as Congress waits to act, the proliferation of state-level laws makes it increasingly difficult to build support for federal bills. As more states pass their own laws, legislators face mounting pressure to defend their state’s particular approach, increasing political opposition to a federal solution.
This paper will first consider the impact that AI is poised to have on the American economy and the state of AI regulation across the country today. Then, it looks back to the development of privacy legislation to examine how our previous failures can help us avoid making the same mistakes as we move to regulate AI.
