It’s been exactly two weeks since the United States Senate voted 99-1 to remove the AI moratorium from the reconciliation package. Overall, that decision was a mistake and a clear setback for technological freedom and digital expression—but the vote was also not a reflection overwhelming disapproval. It is better viewed as a decision to revisit the issue later. Recent reporting suggests Congress will try again. What that approach will look like, when it will happen, who will lead it, and what exactly its effects will be on the computational freedom of consumers are all still uncertain. As the debate re-emerges, the American Consumer Institute will continue to provide thought leadership on the appropriate path forward. In anticipation of the debate’s future reignition, here are some broad takeaways from the proposed AI moratorium that policymakers and policy wonks should take to heart in advance of the next attempt.
1. Small states think they have more power than they do.
A bipartisan group of 260 state lawmakers sent a letter to Congress making the case against the AI moratorium, reminding Congress that they are the “laboratories of democracy” and that a pause “would freeze policy innovation.” None of the arguments were surprising (of course states would fight to protect their turf!). But the signature composition was unexpected: only 4 signatures were from California, and 5 were from New York. Meanwhile, Montana matched New York in number, 15 Idaho lawmakers signed on, and a massive 20 lawmakers from Vermont jumped on board. Utah’s delegation felt so strongly that they sent their own letter of disapproval to the Capitol.
Unfortunately, the loudest voices against the moratorium—including states like Utah, which has done an excellent job of state AI policymaking—will be silenced by California’s geographical proximity and overbearing size as it decides AI policy for everyone. Nearly 12 percent of Americans live in California, and with 32 of the top 50 AI companies within their borders, what politicians in Sacramento say on AI ultimately goes for everyone. That is, unless Congress says otherwise.
This extraterritorial policymaking has become broadly known as the “Sacramento Effect,” and California is showing no signs of slowing down. On top of the 18 AI dictums California signed into law in 2024, it is considering many new proposals to regulate AI this year. With an eye to the future, California’s AI Policy Blueprint report shows how California is already beginning to plan 2026’s legislative push to flex its global regulatory muscle. The report’s recommendations are relatively tame compared to previous efforts, but it still includes language that clearly expresses a desire to “shape the world” while “influencing global standards” that California thinks will create “worldwide impact.”
The reality is that big states are intentionally governing extraterritorially with the mistaken stamp of approval from smaller states that think they hold more sway over AI’s trajectory than they actually do. When it comes to clear matters of interstate commerce and speech, where much larger states can bully smaller states into implicitly adopting their frameworks, small states would be much better represented by working with federal lawmakers in a legislative domain designed to guarantee their equal representation: the United States Senate.
2. Lawmakers seek a pound of Big Tech flesh at any cost—even if it hurts little tech.
Politicians want their pound of flesh from Big Tech, and it ultimately fueled the AI moratorium’s demise. In the void of federal leadership, state laws will continue to pursue their own regulations at a significant cost to businesses and consumers—costs which will grow as other states and jurisdictions add their own unique requirements. As ACI research shows, even when states tried to harmonize their rules around similar AI regulatory concepts and language, they have failed miserably. The proposals and laws are very different as states cannot even agree on how to define key terms like artificial intelligence, high-risk, and consequential decision. Little tech will be left to shoulder the financial burden of complex compliance regimes as Congressional wavering allows differences to proliferate.
Aggressive approaches from California and New York will continue to set baseline rules for everyone. As other states fully bring their AI policy regimes online, the patchwork of costly rules and regulations will deepen. Unfortunately, the true cost of overlayed, multi-jurisdiction regulations has been underestimated in past calculations—particularly in the privacy debates. Of course, large technology companies would prefer a world where they do not need to comply with regulatory chaos, but they will also benefit from the regulatory moat that the mishmash of rules and regulations creates.
Fortunately for Big Tech, they will be saved by their deep pockets, which can employ teams of lawyers, lobbyists, and compliance officers to fight against these laws in the states and then transition their business models and operations into easier compliance efforts when those efforts fail. How fledgling AI firms with thin profit margins are supposed to comply with so many rules and regulations, each with their own unique compliance and reporting requirements, remains much less clear.
3. Union realignment is a future threat to computational commerce.
Here’s how the unions spun the AI moratorium. Teamsters president Sean O’Brien (yes, the same Sean O’Brien that spoke at the Republican National Convention) argued in Fox News that the AI moratorium would create a dystopian vision of the world where Big Tech is given a “decade-long free pass…to kill jobs, silence state governments, and reshape the country to their liking behind closed doors.” The AFL-CIO opposed the AI moratorium as well—suggesting that the provision was a giveaway to Big Tech. Several unions also signed on to a letter opposing the AI moratorium.
If union wishes lead to a tsunami of regulations that ultimately smothers the next technological revolution in its crib, ironically, they will have done so while insulating Big Tech from Little Tech competition. They will have also advanced policy that undermines the domestic industry that employs 94 percent of non-unionized American private sector workers.
As conservatives consider realigning themselves on some positions in ways that they perceive to be more pro-worker and union-friendly, they must grapple with the fact that unions are arguing for policies that benefit certain union members specifically at the expense of all workers generally. Not all regulations are created equal, but on the whole, regulations harm job creation, slow productivity, lower wages, and discourage entrepreneurship. That’s hardly a pro-worker agenda. Of course, AI will lead to certain job dislocations, but when unions sell you a story that more regulations (especially more AI regulations) lead to more jobs than it destroys, don’t believe them. The evidence is not on their side.
4. Existential debates at the heart of AI and family policy
The debates about the AI moratorium are more than economic: they directly wrestle with the role of humans in a world of rapidly advancing technology. Over on his Medium blog, R Street Institute senior fellow Adam Thierer recounts how one of his Congressional testimonies devolved from a discussion on how AI could improve government efficiency into a quasi-religious fight on whether technology companies were trying to “replace God with a new AI deity.” Those concerns reflect questions about whether humans are trying to construct what Jason Hausenloy calls a “modern Tower of Babel disconnected from biblical values.” Objections to certain AI legislation varies by person, within broader classifications, and across many dimensions, but a central component of this debate is to what extent humans are distinct from machines.
Parallel to this debate were concerns from other religious and family groups that separately criticized the policy on the grounds of ethical misuse rather than literal deification. For example, when the AI moratorium made its way into the reconciliation package, the National Association of Evangelicals (NAE) signed onto a letter opposing the moratorium, arguing that it could “silence communities and conscience-driven policymakers working to protect children, families and workers from the harmful impacts of unregulated AI.” Similarly, other groups worried that moratorium would evaporate or prohibit protections for kids and workers.
These kinds of reservations are nothing new—but so far, technological advances have not always played out the way many skeptics predicted. Although not a religious debate, the Luddite movement in England famously thought machines would replace humans with secondary concerns that early mechanization would erode the value society places on human work. Instead, humans found different meaningful jobs. Similarly, the ATM augmented bank teller jobs instead of replacing them. On a much broader level, there is also a long history of technological advancement being seen as tampering with God’s creation, including for many commonly accepted technologies and processes today, like organ transplantation. Of course, artificial intelligence is an unprecedented technology that differs from past technologies in many important respects—but every technology was, at one point, new and the techno-panic predictions that accompany technological change are often misguided because the future is difficult to predict. We should not be cavalier or naïve about the power and risk of AI technology, but we should also have enough humility to recognize that the worst technological fears rarely pan out as anticipated.
5. Congressional dysfunction is making things worse for everyone
The ghost of Congressional dysfunction only supports a continued patchwork. In one opposition letter, state policymakers argued that there has been a “long absence of federal action to address privacy and social media harms.” Key Senate decisionmakers, Marsha Blackburn (R-TN), made the point even more clear: “Until Congress passes federally preemptive legislation like the Kids Online Safety Act and an online privacy framework, we can’t block states from making laws that protect their citizens.” The critics have a point. Congress has done little to address technology policy.
The reality remains that, despite broad agreement inside the beltway that something must be done on data privacy regulation, Congress has decided against legislation time and again. In the view of state policymakers, that gives Washington very little credibility to follow through on commitments to create a federal AI framework. Had Congress passed federal data privacy legislation before the debate on the AI moratorium, it may have bought federal policymakers desperately needed political capital and signaled to the states that it can deliver on broadly supported technology policy ideas. For this reason, as specific AI harms arise, Congress should address them quickly, sending a message to the states that it is willing and able to respond to emerging AI challenges and harms. Fortunately, Congress is showing early signs, such as the Take It Down Act, that it will be able to address some harms in AI that it could not otherwise agree to on social media or privacy policy.
6. Chaos reigns
Congress has abdicated its leadership on technology policy, opening the door to a broad anti-tech backlash that ultimately killed the first try for an AI moratorium in Congress. Now that the dust has settled on the AI moratorium’s demise, chaos still reigns. More than 1,000 proposals—some sensible, others fatally flawed—still fester in America’s laboratories of democracy, and it’s all going to get worse before it gets better.
Although some laws have more clearly anticipated unintended consequences than others, the sheer volume makes it difficult for technological companies of all sizes to comply with an emerging labyrinth of rules, hurting innovation and consumer choice, while allowing certain consequential policy decisions to fly under the radar.
In the end, the AI moratorium fell victim to a realignment of organized interests. But not all hope is lost: the optics of the overwhelming 99-1 vote against the idea looks much worse than reality. Even those who strongly supported the idea decided to move on. Congress should revisit soon. Americans cannot afford to wait much longer.
Logan Kolas is the Director of Technology Policy at the American Consumer Institute, a nonprofit education and research organization. For more information about the Institute, visit us at www.TheAmericanConsumer.Org or follow us on X @ConsumerPal.