The Blackburn-Cruz AI Regulatory Pause Compromise Offers a Path Forward
On June 30, Sens. Marsha Blackburn (R-Tenn.) and Ted Cruz (R-Texas) announced a major compromise in the debate over a proposed moratorium on state and local AI-specific regulatory enactments. The senators, who had been at odds over the moratorium’s scope, agreed to new language that shortened the length of the pause from 10 years to five years and added further clarification on what is exempt from the proposal. A different version of the AI pause already passed in the House as part of the “One Big Beautiful Bill Act.” While the Senate bill took a slightly different approach, it also included a 10-year pause on state and local AI regulatory enactments.
The debate in the Senate proved contentious, and Blackburn emerged as a leading critic of the AI regulatory pause, necessitating a compromise with Cruz and other backers of the idea. In addition to winning a shorter length of time for the duration of the regulatory pause, Blackburn was able to achieve specific concessions about it not applying to “a law or regulation pertaining to unfair or deceptive acts or practices, child online safety, child sexual abuse material, [and] rights of publicity.”
Essentially, this new language clarifies what was already true—namely, that the moratorium exempts “generally applicable law” from its coverage. Generally applicable laws include a broad category of policies like unfair and deceptive practices law, civil rights law, product recall authority, court-based common law remedies, and a variety of other consumer protections. The Blackburn-Cruz compromise simply adds a few specific mentions of particular types of generally applicable laws that would be exempt from the moratorium.
However, some critics are still debating what the closing language of this provision means when it caveats the new clarifications by noting that the compromise exempts state laws so long as they do not impose an “undue or disproportionate burden” in order to “reasonably effectuate the broader underlying purposes of the law or regulation.” Critics worry that this language could be used to undercut the new clarifications or even to undermine the notion that the moratorium does not apply to generally applicable law.
I do not read it that way. While some opponents may try to scare people into thinking that big loopholes exist in order to stop the regulatory pause proposal, the precise opposite could happen: State and local governments will likely find creative ways to contort the “generally applicable law” exception to push through many laws that still regulate AI without calling it such.
To be clear, even with the new compromises, this remains a much-needed measure that will help address the growing patchwork of over 1,000 AI-related laws popping up across the nation. Congress must take action to protect the market for interstate algorithmic commerce and constrain state-by-state overregulation before it undermines America’s global competitiveness and makes it harder for U.S. firms to remain competitive with China. Even if the AI regulatory moratorium lasts only five years, that is still a good start. Congress can continue to work with states to determine a sensible approach to AI policy in the coming years. A recent R Street analysis also discussed several specific policy initiatives Congress can take up immediately as lawmakers continue to consider AI governance.
Nonetheless, many challenging questions remain about how the AI regulatory moratorium will be enforced. What we are learning from the political fight over the AI regulatory pause is that legislative drafting is complicated and open to considerable interpretation. Lawmakers are confronted with complex trade-offs when deciding whether to speak in broad terms or get hyper-specific about provisions and exemptions. Taking the latter route sounds smart, but it can quickly become a contentious political spectacle with various interests seeking an ever-expanding list of exemptions. Practically speaking, that approach often fails to accomplish anything due to protracted political fights.
The new Blackburn-Cruz language strikes a reasonable compromise by speaking in general terms about the scope of the AI moratorium’s coverage while offering targeted verbiage on a few specific issues where Blackburn and others requested greater clarification. This approach offers a path forward to ensure Congress’s voice is heard and can help shape policy for the national AI marketplace before a patchwork of costly, confusing regulations decimates innovation, investment, and competition in this vital technology sector.