In the Race Between the Federal Government and the States to Regulate AI, Big Tech is the Winner
Matthew Berger
As artificial intelligence (“AI”) technology and capabilities expand, a race is being set up between the federal government and state legislatures to regulate it. In the 2025 legislative session, all fifty states (plus Puerto Rico, the Virgin Islands, and Washington, D.C.) introduced some manner of AI legislation. The federal government has taken a different approach, hoping to win the AI race through preemption and deregulation. Running alongside them is Big Tech, a collective term for the most influential technology corporations in the world (Alphabet, Amazon, Apple, Meta, Microsoft, etc.). These companies are expected to spend more than $300 billion on AI in 2025 and are working to recoup those costs through regulations favorable to their pro‑innovation agenda.
Part I of this Comment provides a general background on AI technology before exploring the federal and state-level approaches to AI regulation. It will also look at Texas Senator Ted Cruz’s proposal for an AI regulatory sandbox in the United States, as well as similar sandbox programs in the European Union. Part II will then analyze weaknesses in the federal and state proposals, arguing that a federal sandbox program will result in tech companies getting a free pass to avoid federal regulation. Part II also argues that shifting the focus at the state level to transparency over liability for catastrophic damages will do little to hold tech companies accountable and internalize harm. Given the weaknesses in current federal and state legislation, this Comment argues that state-led regulation that holds AI developers and deployers strictly liable for abnormally dangerous AI activity could be an effective way to keep Big Tech accountable and protect consumers. Finally, recognizing that a liability-based approach will have its own weaknesses, this Comment will offer the Dynamic Governance Model as a means of combining federal and state approaches and working towards collaborative AI governance.
While collaboration would be ideal, the current regulatory course is set up for a race between federal deregulation and state-level transparency. Without a liability system to act as a referee, however, the winner of this race will be Big Tech.