HB Ad Slot
HB Mobile Ad Slot
US State AI Legislation: Virginia Vetoes, Colorado (Re)Considers, and Texas Transforms
Friday, March 28, 2025

Virginia’s Governor, Glenn Youngkin, vetoed a bill this week that would have regulated “high-risk” artificial intelligence systems. HB 2094, which narrowly passed the state legislature, aimed to implement regulatory measures akin to those established by last year’s Colorado AI Act. At the same time, Colorado’s AI Impact Task Force issued concerns about the Colorado law, which may thus undergo modifications before its February 2026 effective date. And in Texas, a proposed Texas Responsible AI Governance Act was recently modified.

The Virginia law, like the Colorado Act, would have imposed various obligations on companies involved in the creation or deployment of high-risk AI systems that influence significant decisions about individuals in areas such as employment, lending, health care, housing, and insurance. These obligations included conducting impact assessments, keeping detailed technical documentation, adopting risk management protocols, and offering individuals the chance to review negative decisions made by AI systems. Companies would have also needed to implement safeguards against algorithmic discrimination. Youngkin, like Colorado’s Governor Polisworried that HB 2094 would stifle the AI industry and Virginia’s economic growth. He also noted that existing laws related to discrimination, privacy, data usage, and defamation could be used to protect the public from potential AI-related harms. Whereas Polis ultimately signed the Colorado law, Youngkin did not.

However, even though Polis signed the Colorado law last year, he urged in his statement for legislators to assess and provide additional clarity and revisions to the AI law. And, last month, the AI Task Force issued a report on their recommendations. The task force identified potential areas where the law could be clarified or improved. It divided them into four categories: (1) where consensus exists about changes to be made; (2) where consensus needs additional time and stakeholder engagement; (3) where consensus depends on resolving multiple interconnected issues; and (4) where there is “firm disagreement.” In the first are only a handful of relatively minor changes. In the second, for example, is clarifying the definition of what are “consequential decisions” – important because AI tools used to make them are the ones that are subject to the law. In the third, for example, is defining “algorithmic discrimination” and obligations developers and deployers should have in preventing it. And in the fourth, by way of example, is whether or not to include an opportunity to cure incidents of non-compliance.

Texas, like Colorado and Virginia, has been considering legislation that addresses high-risk AI systems that are a “substantial factor” in consequential decisions about people’s lives. That bill was recently modified to remove the concept of algorithmic discrimination, and as currently drafted prohibits AI systems that are developed or deployed with the “intent to discriminate.” It has also been modified to expressly state that disparate impact alone is not sufficient to prove that there was an intent to discriminate. The proposed Texas law is similar to Utah’s AI legislation (which went into effect on May 1, 2024), insofar as it would require notice if individuals were interaction with AI (though this obligation is only for government agencies.) Lastly, the law would also prohibit the intentional development of AI systems to “incite harm or criminality.” The law was filed on March 14 and, as of this writing was pending in the House Committee.

Putting it into Practice: The veto of HB 2094 emphasizes the complex journey towards comprehensive AI regulation at the state level. We anticipate ongoing action at a state level as legislatures, and some time before we see a consensus approach to AI governance. As a reminder, there are currently AI laws in effect focusing on various aspects of AI in New York (likenesses and employment), California (several different topics), Illinois (employment), and Tennessee (likenesses), passed AI legislation set to go into effect at different times in 2024 through 2026, and bills sitting in committee in at least 17 states. 

Listen to this post 

HTML Embed Code
HB Ad Slot
HB Ad Slot
HB Mobile Ad Slot
HB Ad Slot
HB Mobile Ad Slot
 
NLR Logo
We collaborate with the world's leading lawyers to deliver news tailored for you. Sign Up for any (or all) of our 25+ Newsletters.

 

Sign Up for any (or all) of our 25+ Newsletters