What would have been a major turning point for tech governance, a moratorium on state-level artificial intelligence regulations, collapsed during the 27-hour Senate “vote-a-rama” amendment process to President Donald Trump’s budget, tax, and immigration plan. With Trump’s July 4 signature making it law, policymakers across the country are scrambling to reassess their approaches to AI oversight.

Meanwhile, with the state-level AI regulation ban left out of the legislation, a domestic policy cornerstone in Trump’s second, nonconsecutive term, advocates of a moratorium are regrouping. They’re working up plans to make another push in Congress. In the meantime, the future of AI regulation remains uncertain.

Recommended Stories

• Tyrants need buffers: China is not a monolith. That's an opportunity for the West.
• Turns out that, for Trump, breaking up with Putin is hard to do
• Endangering Article III: The Supreme Court may not have done enough to prevent a reckoning

The House version of what Trump called his “one big, beautiful bill” contained a 10-year moratorium on state AI regulation tied to internal government spending on AI. A Senate version instead tied the legislative pause to states’ eligibility for federal funds under a program from former President Joe Biden’s administration, the Biden-era Broadband, Equity, Access and Deployment program, intended to connect more underserved areas to the internet. The parameters of the moratorium were also clarified and narrowed in an effort to gain support in the Senate. 

After a reported compromise, reducing the duration of the moratorium from 10 to five years, was reached by a supporter, Sen. Ted Cruz (R-TX), and an opponent, Sen. Marsha Blackburn (R-TN), the latter abruptly moved to cut the provision from the bill, with her amendment passing in a 99-1 vote. Later, in a statement, Blackburn said of the moratorium that the “provision could allow Big Tech to continue to exploit kids, creators, and conservatives, and I am pleased it will no longer be included in the One Big Beautiful Bill Act.”

Support for the moratorium came from the tech industry and antiregulatory groups, but opposition from labor groups, state officials, and child privacy advocates carried the day. The moratorium was not part of the final bill that Trump signed into law at an Independence Day ceremony at the White House.

Advocates of the pause in state action voiced concerns that complying with a patchwork of different, and sometimes contradictory, state regulatory regimes would place U.S. developers at a disadvantage to their counterparts in China. They also argued that federal regulatory action is more appropriate for AI because those products and services are interstate commerce, not technologies generally confined to a state’s border. One national policy also avoids favoring the same large U.S. tech companies the government is prosecuting as monopolists in various antitrust suits, as nascent competitors would not have the same resources to comply with a state patchwork of regulatory regimes.

But opponents of the moratorium were pleased with its absence in the final version of the bill. Brad Carson, the president of Americans for Responsible Innovation, a proregulatory nonprofit organization, said in a statement, “Let this be a lesson to Congress — freezing state AI laws without a serious replacement is a political nonstarter.”

That replacement of a federal AI regulatory framework might be what’s next for Congress to debate. While some states, such as Montana, Ohio, and New Hampshire, are moving toward “Right to Compute” measures to protect activities essential to AI development, there are hundreds of state bills and laws widely understood to restrict AI activities across the country. If Congress wants to assert its authority or halt those state efforts, it will have to do so proactively with its plan, perhaps involving state preemption.

Still, others warn against regulating AI at either the state or federal level. Jim Harper of the American Enterprise Institute reacted to the failure to pass the state moratorium with concerns about federal regulation of AI, too. He blogged, “Nobody – including at the federal level – actually knows what they are trying to regulate.” Referring to the widespread prevalence of computer-aided decision-making across applications, services, and industries, he wrote, “Defining AI is fraught because ‘artificial intelligence’ is a marketing term, not a technology.” Harper advocates instead regulating bad outcomes, no matter what technology was used to achieve them.

SENATE GOP SADDLED WITH SIGNALGATE AS WALTZ PLUCKED TO FILL LAST CABINET POST

It remains unclear if Congress will take such a principled approach in crafting a regulatory AI framework, but it’s sure to risk a redux of its inability to pass a federal privacy framework. Congress has been stymied in its attempts to pass comprehensive privacy legislation by politically driven conflicts around a private right of action and preemption. Now, a state regulatory patchwork, with attendant compliance costs, has materialized across the nation.

Congress risks a similar legislative quagmire as it tries to move forward on AI policy, all with midterm elections just around the political corner.