US State AI Laws Surge as Federal Regulatory Vacuum Persists
A wave of state-level legislation targeting chatbot safety and deepfakes is creating a fragmented compliance landscape for AI developers.
US state legislatures are rapidly filling a federal regulatory void with a surge of AI-specific laws. This shift marks a transition from introductory policy discussions to the enactment of strict mandates regarding chatbot safety, data provenance, and synthetic media disclosures.
The scale of this legislative movement is accelerating. In 2025, US state legislatures enacted 145 AI-related bills, representing a 47% increase over the 99 bills passed in 2024. This momentum intensified in early 2026; by the end of the first quarter, 45 states had introduced 1,561 AI bills. This volume exceeded the total number of AI-related bills introduced across all states for the entirety of 2024 within just three months.
A Patchwork of State Mandates
In the absence of comprehensive federal guidance, individual states are implementing divergent approaches to AI risk. Vermont has taken one of the most restrictive stances, with Governor Phil Scott signing law H.816, which bans the use of AI chatbots to provide therapy.
Meanwhile, Arizona has focused on a broader set of consumer protections. The state passed three AI-related bills in 2026, including a specific measure for chatbot safety and an expanded law requiring the disclosure of synthetic images, commonly known as deepfakes. These actions highlight a growing trend toward mandatory transparency and the legal requirement to distinguish between human and machine-generated content.
The Compliance Burden
This fragmented regulatory environment creates significant operational hurdles for AI developers. Rather than adhering to a single national standard, companies must now navigate a patchwork of 50 different sets of rules. The increasing focus on data provenance—the tracking of where AI training data originates—suggests that developers will soon face stricter accountability for the datasets used to build their models.
For the industry, this means that a feature legal in one state could be prohibited in another, complicating the deployment of automated decision-making tools and consumer-facing bots. The shift toward state-led regulation effectively turns the US into a laboratory for AI policy, where the most restrictive states may set the de facto national standard for corporate compliance.
Future Outlook
As states continue to prioritize consumer protection against deceptive AI interactions, the industry is watching for whether these disparate laws will eventually coalesce into a unified federal framework. For now, the focus remains on the increasing granularity of state laws, particularly those targeting high-risk sectors like healthcare and digital identity. Whether federal legislators will act to preempt this state-level fragmentation remains the primary uncertainty for the sector.