Why Adversarial AI Laws Are the Wrong Move for America 

The race to develop Artificial Intelligence (AI) isn’t just a race against our foreign adversaries. It’s also a race against the country’s policymakers who are rushing to write legislation before they even understand what AI means, how it’s applied, and how it might be regulated without upending the economy. Amid this conversation about what to do, lawmakers often overlook the sheer breadth of AI’s impact, spanning everything from predicting diseases to powering productivity tools like ChatGPT. AI is not a standalone technology but a foundational element across a wide range of consumer services.

Take Netflix, for example. It’s not considered an “AI company”, but the go-to streaming service for many Americans has incorporated an AI-driven recommendation system that shapes everything from what shows are suggested to how thumbnails are displayed. Then take Amazon, the company that puts everything from groceries to gadgets at our fingertips. Every “Recommended for you” suggestion, every voice command on Alexa, and every optimized delivery route is powered by AI. AI lets Amazon fine-tune shipping routes, stock shelves predictively, make useful product recommendations, and – perhaps most importantly – keep consumer costs low. 

Imprecise policies would classify Amazon’s complex logistics and recommendation systems as “AI-driven,” and a nightmare of compliance costs and restrictions would kick in. The result would be slower delivery times, more “out-of-stock” alerts, and a very different Amazon than consumers expect. Small businesses selling on Amazon’s marketplace also would feel it, facing unpredictability. Likewise, if lawmakers forced Netflix to ditch its recommendation tools or jump through legal hoops to keep them, not only would the company suffer, but the consumers would lose out on the services that they’ve grown accustomed to. 

The list of companies that use AI, but not as the core component of their service, goes on and on. At best, we’d lose useful services. At worst, the platform might have no choice but to head for the door.

This is what states like California and Texas are sure to find out. 

California has developed a bad habit of driving out household tech names over the past five years. HP, Oracle, and Tesla didn’t just relocate on a whim, they fled from California’s tightening regulatory stranglehold. Most recently, California almost doubled down, sending a bill, SB 1047, all the way to the Governor’s desk that would have skyrocketed compliance costs for countless California-based companies using AI.

In the past, those same California companies fled to Texas. But now, in an ironic twist, Texas policymakers are flirting with the same style of regulation that sent California’s companies running. The proposed Texas Responsible AI Governance Act (TRAIGA) applies a sweeping regulatory framework that classifies many different kinds of AI applications as “high-risk,” subjecting everyday companies to “prior conformity assessments,” ongoing audits, and penalties up to $100,000.

Companies won’t stick around to learn how to navigate costly new compliance regimes; they’ll pack up and leave for states where AI isn’t a dirty word. Moving forward, the real question is where these companies go if not even Texas will welcome them.

As companies begin their search for a new home, pro-growth states like Utah and Florida have the greatest opportunity to steal away America’s most valuable economic contributors. While other states like Colorado fail to recognize the opportunity, putting up walls through bills with overly broad definitions of AI, Utah and Florida could lead the way by passing legislation that ensures “AI” isn’t a dirty word.

So far, both states have demonstrated that effective AI legislation requires a scalpel, not a sledgehammer. Utah’s recent legislation, SB 149, created a three-year AI Innovation Lab, designed to work with innovators to learn, above all, about the facets of AI and how regulatory proposals would affect it. Moreover, when Florida confronted the challenge of AI-generated political content it chose to require disclaimers of AI use rather than imposing outright bans. To ensure future regulatory measures are knowledge-driven rather than reactive, Florida established a dedicated council to provide input on emerging issues. 

By continuing to recognize the importance of precision in legislation, these states can avoid the pitfalls driving billions in GDP away from Silicon Valley – and soon – from Austin. But if states are serious about capitalizing on the AI revolution, they need to do more than just avoid bad policy – they need to build a foundation that actively welcomes innovation while working hand-in-hand with industry to develop the infrastructure that supports growth. 

For starters, this means streamlining permits for data center construction, embracing market-driven energy solutions that provide reliable power for compute-intensive AI operations, and creating flexible frameworks to allow autonomous vehicles to test and deploy without excessive red tape. States that embrace these practical, pro-market steps today will become the tech hubs of tomorrow.

This article was authored by Devin McCormick, Tech & Innovation Policy Analyst at Libertas Institute, and Turner Loesel, Technology & Innovation Policy Analyst at the James Madison Institute, based in Tallahassee, Florida.