Despite the heated national debate on AI, the biggest AI legislation fights of 2024 happened in statehouses, not Congress. Now, the House Energy and Commerce Committee aims to take control of AI policy at a national level with a new measure in a must-pass budget reconciliation bill: a moratorium on state-level AI legislation that cements Congress as the sole authority on AI. And with the measure, the startups poised to take advantage of the growing AI market have the chance to cement themselves as a political force, not just a market one.

Last summer, a coalition of startup founders, investors, and new media played a significant role in killing California’s onerous SB 1047, which would have created a new state agency to regulate AI — and which critics argued would “stifle innovation in the AI sector and hurt California’s economy” — through intense, dogged engagement with politicians and media. Now, the Energy and Commerce Committee’s moratorium gives this same group a key opportunity to prove they’re a lasting constituency with clout.

The most ambitious attempts to regulate AI to date have come from state-level bills like SB 1047. Just as California’s Clean Air Act waiver allows the state to dictate emissions policy for the whole country without going through the federal legislative process, state AI bills would effectively regulate all American artificial intelligence companies.

Congress is getting tired of large states trying to bypass it. The moratorium, which recently passed the House Energy and Commerce Committee, would make Washington the exclusive regulator of AI. Representative Jay Obernolte (R-CA23) articulated the logic behind the measure: “No one believes that AI should be unregulated… but it should be the job of Congress, not the states.” Crucially, the measure is constitutional. The Interstate Commerce Clause — a provision in the U.S. Constitution — authorizes federal regulation of economic activity that crosses state lines. Congress can invoke it to argue that only the federal government can set rules for industries operating across state lines, which includes nearly all major tech companies and the AI startup world.

By organizing behind the moratorium, the same coalition that played a pivotal role in killing SB 1047 can show Congress there’s a new, well-organized interest on the block — and help lock in a meaningful shift toward growth-focused governance.

In 2024, national politicians of both parties distanced themselves from SB 1047. A Pelosi-backed letter from House Science Democrats criticized the bill for having “little to no basis in evidence.” Since then, bipartisan deals struck in the House and Senate have avoided major AI legislation. And now, the most important AI issue for both parties has become staying ahead of China.

Consequently, supporters of aggressive AI regulation have turned to state legislatures, helping to draft over 1,000 AI-related laws. So far, most ambitious ones have failed. SB 1047 was vetoed by Governor Newsom, while TRIAGA, another controversial state bill out of Texas, only got through the statehouse after it was revised to impose minimal requirements on the private sector and focus instead on government use of AI (it’s currently under consideration in the Texas Senate). The broadest AI law passed in 2024 — SB 205, a “comprehensive” AI law similar to the original TRIAGA — was signed into law in May 2024 by Colorado Governor Jared Polis, but doesn’t go into effect until 2026. Polis, a Democrat, now supports the national AI moratorium, which would supersede the regulations imposed by SB 205. “He is generally supportive of a federal moratorium to give Congress a chance to figure this out and create a true 50-state solution to smart AI protections,” a spokesman for Polis told Bloomberg.

While the moratorium passed committee on a party-line vote, it’s scrambling political coalitions elsewhere. “It says a lot that the one governor whose state has enacted ‘comprehensive’ AI regulation supports a moratorium on state-based AI regulation,” Dean Ball, a Senior Policy Advisor for AI and Emerging Technology at OSTP, posted on X.

The over 1,000 state AI bills have already caused confusion for AI founders, particularly at startups, which lack the policy teams to comply with regulatory uncertainty. “Innovation and investment require clarity and certainty, and individual states creating a patchwork of arbitrary requirements to fit their local political agenda will just hold open source and American AI back” Nous Research CEO Jeffrey Quesnelle told me. “If / when there’s AI regulation, it needs to at least be clear, consistent, and uniform.”

“Every extra bit of friction reduces a startup’s chance of winning,” said Bayeslord, a pseudonymous startup founder and co-founder of the e/acc movement. “Startups deploying AI need 50 different state AI compliance regimes like they need a hole in the head,” he added.

Regulatory barriers tend to hit startups hardest because they lack the funds to lobby for policies or often even to comply with them. As a consequence, incumbents often attempt to pass regulations that increase barriers to entry for their competitors, a process known as regulatory capture. “Early bans on emerging technologies often failed to reduce harm and instead left restrictive regions behind, both as producers and consumers, as innovation advanced elsewhere,” venture capitalist and author Steven Sinofsky told me.

The startup coalition and its allies in new media have an opportunity to change that. By rallying tightly connected networks of startup founders, academics, and influencers, the group successfully killed SB 1047. Now, the next major test has come, and whether they can make themselves heard on the national stage could make or break its future.

—Brian Chau

Feature image — Midjourney user karinrosenberget