Skip to main content

3 Trends Emerge as AI Legislation Gains Momentum

Consumer protection, deepfakes and government use of AI drew state lawmakers’ attention in 2024.

By Chelsea Canada  |  January 23, 2025

Lawmakers in 2024 proposed a variety of artificial intelligence-related bills in most U.S. states, Puerto Rico, the Virgin Islands and Washington, D.C. NCSL tracked over 450 bills in 23 different AI-related categories and found three legislative trends rising to the top: consumer protection, deepfakes and government use of AI.

A couple of states passed first-in-the-nation AI legislation focused on consumer protections. At least half the states addressed deepfakes through new laws targeting the technology’s use in elections and sexually explicit materials. Finally, most states considered or enacted bills related to government use of AI tools.

Consumer Protection and Transparency

Lawmakers considered over 100 bills in the two categories of private sector and responsible use. Three states passed the first U.S. laws focused on safety and protections for consumers when using AI products.

Colorado passed the nation’s first comprehensive AI law (SB 205). The new law applies to AI systems that can impact consumers in consequential ways in education, employment, financial or lending services, government, health care, housing and more. The law says AI developers and deployers must avoid algorithmic discrimination, defined as any use of AI that results in unlawful differential treatment or that disfavors a group of individuals protected under current state and federal laws based on age, disability, ethnicity or other protected class. Entities must also meet certain requirements that focus on consumer protections, risk management and transparency. Another new law (H 1468) created an AI task force to recommend changes to SB 205 prior to its implementation in 2026.

A new Utah law (SB 149) requires specific entities to disclose use of generative AI. For violations, the Division of Consumer Protection can impose fines up to $2,500. The division may go to court for an injunction to impose other fines or to recoup attorney fees and court costs. The attorney general also may seek $5,000 in civil penalties for each violation of an administrative or court order tied to the new law.

California Gov. Gavin Newsom vetoed a bill (SB 1047) that would have added new requirements to the development of large AI models by setting out various testing, safety and enforcement standards. But the state enacted a different law (AB 2013) that focuses on training data transparency. Starting in 2026, developers of generative AI systems must publicly share what data was used to train the system or service.

Deepfakes

At least half of states passed over 40 new laws for deepfake technology, which uses AI to create false audio, images or video. At least 19 states passed legislation related to sexually explicit deepfakes. Of those states, 12 focused generally on fabricated sexually explicit images, and others homed in on these types of materials that depict minors. Florida enacted legislation (Title XLVI, Chapter 827) to establish a crime for the creation of computer-generated child pornography. Washington’s new law (H 1999) amends the state’s child pornography law to include digitally fabricated content. Among other elements, the law creates a civil cause of action for adult and minor victims impacted by the nonconsensual disclosure of artificial intimate images. Indiana (H 1047) expanded the state’s revenge porn law to include unauthorized AI-generated content.

Over 10 states enacted laws related to the use of deepfakes in election and political messaging. Arizona passed two election-focused AI bills in 2024. One new law (H 2394) allows for any candidate to sue if a “digital impersonation” of the person is published. The second new law (S 1359) requires disclosure of the use of a deepfake of a political candidate within 90 days of an election. Utah (S 131) also enacted a new law requiring disclosure of a deepfake in a political advertisement for a candidate or ballot measure, but the law applies any time before a primary or an election. (For more information about deepfakes in election and campaign materials, see the NCSL brief AI in Elections: A Look at the Federal and State Legislative Landscape.)

Government Use

State legislators considered over 150 bills relating to government use of AI in 2024. Although new regulations are being discussed and rolled out, state agencies are already using AI to improve physical infrastructure, optimize government resources and assist citizens with inquiries.

At least 10 states, including Connecticut, Delaware, Maryland, Vermont and West Virginia, directed state agencies to inventory and describe AI applications within their operations that impact the services they deliver. To address concerns about possible bias, discrimination and disparate impact, states such as Connecticut, Maryland, Vermont, Virginia and Washington mandated that state agencies perform impact assessments to ensure that the AI systems in use are ethical, trustworthy and beneficial.

As state legislatures kick off their 2025 legislative sessions this month, work on many of these AI issues will continue as interest increases on the impacts the technology can have. For more information on AI legislation and additional resources on AI, visit NCSL’s AI Policy Toolkit.

Chelsea Canada is a program principal in NCSL’s Fiscal, Telecom and Banking Program.

Loading
  • Contact NCSL

  • For more information on this topic, use this form to reach NCSL staff.