Why AI Policy Thrives in Some States and Fades in Others
Discussions on AI are occurring at all levels of government including federal, state, and local as each entity tries to understand, leverage, and protect from AI. In August 2025, states leading and lagging in AI governance were identified, alongside the salience of different issues associated with the technology. This analysis unpacks the factors that lead to the introduction of bills at the state level and the constraints impacting these efforts.
Some states are racing ahead with detailed frameworks, while others have only introduced a few bills. This patchwork emergence of AI legislation shapes how citizens are protected from abuse, addresses AI-generated deepfakes and automated decision-making, and influences the establishment of institutions capable of governing AI over the long term. It also provides context for the current efforts by the White House to curb state-led legislation.
Legislative Activity Analysis
This study examines all AI-related bills introduced from January 2023 through October 2025 across all states to understand which are acting, which are not, and why. A total of 385 bills were tracked, with at least one bill introduced in all 50 U.S. states. These bills were grouped into three overarching themes:
- Protection of the individual
- Transparency and trust in information ecosystems
- Responsible systemic governance
We then linked legislative activity in each state to a small set of structural and political conditions, including:
- Population Age Structure: Older populations tend to correlate with lower legislative activity.
- Per Capita Income and Poverty: Higher incomes and lower poverty levels indicate greater success in lifting residents above the poverty line.
- Governor Party Affiliation: Democratic governors correlate with higher legislative activity.
High and Low Activity Configurations
Our analysis identified patterns in states with high and low activity regarding AI legislation. High-performing configurations emerged primarily in Democrat-leaning states with younger populations, such as New York, California, Illinois, and Maryland. In contrast, low-performing configurations were characterized by Republican-leaning electorates, older populations, and lower per capita incomes.
Barriers to AI Governance
Two distinct barriers to AI governance were identified:
- Material Barrier: Limited fiscal and institutional capacity prevents some states from acting even when risks are recognized.
- Ideological Barrier: Regulatory skepticism and market-oriented political preferences constrain action, even where capacity is strong.
These barriers reproduce familiar divides in U.S. policy innovation, echoing long-standing patterns in areas such as environmental regulation and health technology.
Recent Developments
A significant recent development is President Trump’s executive order signed on December 11, 2025, which seeks to preempt state-level AI laws and promote U.S. dominance in artificial intelligence. This order directs the U.S. attorney general to challenge state laws viewed as impediments to AI leadership, potentially consolidating AI authority at the federal level.
The ongoing debate surrounding this issue, especially after a Senate bill to impose a 10-year moratorium on state AI regulation was defeated, indicates that the role of states in crafting AI laws remains unresolved.
