AI Policy Watch: Introducing our Data
Plus: Where Lawmakers from 20+ states and Both Parties (!) Agree AI Should be Checked
Welcome to the first iteration of AI Policy Watch! As discussed previously, each week we will bring you an update on some of the most important recent developments in AI policymaking throughout the country alongside a comprehensive and organized dataset including every AI bill being considered. Let’s get started with an introduction to our data.
AI Legislation Considered in 2025
As of this week we have identified 499 AI-focused bills that have been introduced across the country for consideration in 2025. We also now have had our first AI bill of 2025 become law - New York S 822 will introduce new transparency requirements for AI use in state agencies and add protections for state employees against the potential negative impacts of AI implementation. Broadly speaking, most of the legislation we’ve collected is still in the early stages of its development.
We’ve categorized these bills in two ways. First, we’ve grouped bills that either expressly expand or promote the use of artificial intelligence and bills that place restrictions on the use of artificial intelligence into the buckets below.
Restricting AI - 314 bills
Promoting AI - 47 bills
Clicking into those links will allow you to access the full list of bills and break them down by jurisdiction, sponsors, and status. Plural’s AI Policy Analyzer can also be used to better understand specific bills.
Our second layer of categorization sorts bills into the more specific topics below. Note: Some bills fit into multiple categories and are therefore included in multiple lists.
AI in Education - 57 bills - Most commonly adding requirements for schools to teach about artificial intelligence, like Georgia HB 487.
AI in Elections - 40 bills - Generally relating to restrictions on the use of AI in campaign materials.
AI in Government - 125 bills - Including initiatives to increase government use of AI, protections against government use of AI, and government-led studies of AI.
AI in Healthcare - 57 bills - Regulating the use of AI in healthcare settings.
AI Use Restrictions in the Private Sector - 123 bills - Legislation restricting the use of AI for various purposes in Non-governmental settings, from the pricing of housing to making employment decisions.
Data Centers - 8 bills - Specifically responding to the impact of AI-related data center growth, including on the environment.
Public Funding of AI - 38 Bills - Appropriations for projects involving the use of artificial intelligence.
Regulated Content - 123 bills - Regulating the content that is produced using artificial intelligence, including explicit material, deepfakes, and AI-powered chatbots.
Regulating AI Developers - 44 bills - Legislation including provisions that directly impact developers of AI technology including large language models.
This legislation is ubiquitous throughout the country, but is most common (so far) in highly populated states including New York, Texas, and Illinois.
In successive weeks we will continue to update this data, and we will further analyze how it is changing over time.
Regulating the Use of AI by Health Insurance Carriers
The U.S. health insurance industry uses massive quantities of data to make millions of important decisions; it is unsurprising that companies in this space have looked to take advantage of advancements in artificial intelligence to support that work. But this application of AI has come under increased scrutiny. Reporting has detailed the adverse impacts of AI-driven healthcare decisions, multiple insurers are facing class-action lawsuits over their use of the technology, and California passed landmark legislation restricting insurers’ use of AI.
Now lawmakers in more than 20 other states have their eyes set on (at least to some extent) following California’s lead. Thus far, there even seems to be bipartisan momentum on the issue. Democrats are more likely to sponsor these bills overall, but 11 of the 37 bills we identified that target the use of AI by insurers had at least some Republican support. That includes Indiana SB 480, which limits the use of prior authorization reviews completed using artificial intelligence. The bill is sponsored or co-sponsored by a majority of the Senate, and appears likely to move towards a final vote in the Senate following its passage out of committee last week.
Illinois HB 35 goes further, and would specifically increase regulatory oversight of AI use by insurers while mandating human contributions to adverse decisions made by insurers and reviews of the use of AI models. The February 18 public hearing for HB 35 saw 34 witnesses speak in favor of the measure compared to just 4 opponents.
It is still too early to tell whether the momentum that pushed California’s bill across the finish line last year can be replicated by advocates for similar measures in other states. It is certainly something we will keep an eye on as sessions wear on.
Until next week, thanks for reading!