In late March, Governor Glenn Youngkin vetoed Virginia’s proposed AI fairness legislation. Had it passed, Virginia would have become the second state to pass similar legislation, starting a state patchwork of AI fairness regulations. Instead, Governor Youngkin sent a clear message: Virginia is—and will continue to be—a pro-innovation hub where AI innovators can set their roots, innovate, and flourish. Here are five takeaways for other states to consider as they debate AI fairness proposals of their own:
- Burdensome AI Regulations Harm Consumers
In his veto statement, Governor Youngkin rightly noted that Virginia’s AI Fairness bill would have had dire consequences. He explained that during a successful 2024, 10 thousand startups were launched, “bringing thousands of jobs, businesses, and investing dollars into the state,” and “[Virginia’s AI Fairness bill] would have undermined this progress, ‘turning back the clock’ and stifling AI.”
The policy proposal would have imposed complex reporting and compliance requirements on AI firms at every level of deployment to mitigate algorithmic unfairness risks. This means firms of all sizes, small or large, would have been subject to the same requirements and penalties—costing Virginia innovators an estimated $290 million and disproportionately harming the same startups Virginia just welcomed. After all, small businesses do not have the same legal resources at their disposal as larger incumbents. The result would have been fewer businesses, less competition, and slower innovation. All of that is bad for consumers.
- Discrimination Is Already Unlawful
Each state already has anti-discriminatory laws, all of which build on or default to existing federal protections. The Virginia Human Rights Act, for example, safeguards all individuals in the Commonwealth from unlawful discrimination based on a list of protected characteristics in education, employment, public accommodation, and real estate. Whether through AI, oral communication, or any other means, discrimination would still be unlawful with or without Virginia’s AI Fairness bill. Governor Youngkin was right to recognize that the bill would have harmed consumers and offered little added protection. In the event of novel threats that only AI could introduce, his administration “has set up safeguards and oversight for AI in its taskforce.” States should consider similar approaches to identify and mitigate specific AI harms rather than sweeping regulations for hypothetical risks.
- AI Evolves Faster Than Legislation
Governor Youngkin further observed that the bill’s “rigid framework fails to account for the rapidly evolving and fast-moving nature of the AI industry.” Other states should take note. Data from Epoch AI shows AI’s fast development, as only 4 of 81 large-scale listed models existed before 2021, but even their data predated major AI moments like DeepSeek in late 2024. Meanwhile, slow-moving bills like Virginia’s AI Fairness bill feature delayed effective dates, with Virginia’s bill pushing back enforcement to July 1, 2026. If the fast trends in AI development continue, the industry will be unrecognizable in mid-2026, rendering many current bills outdated or obsolete before they ever take effect. Instead of trying to race AI, policymakers should recognize existing laws can already apply to AI and defer new action until the emergent tech—and appropriate responses to novel, new harms—are understood better. As the American Enterprise Institute’s Will Rinehart put it in The Dispatch, “Sometimes the best solution is the one you already have.”
- A Different Kind of Patchwork Looms
Had Virginia passed its AI fairness proposal, it would have joined Colorado as the second state to pass broad regulation for algorithmic discrimination. Both bills were part of a multi-state effort to coordinate AI policy, but American Consumer Institute research found participant bills were too dissimilar to build a national standard. The resulting patchwork of reporting requirements, carveouts, and enforcement solutions would have significantly stymied innovation. Virginia’s veto and Texas’s improving and resubmitting its proposed bill signal some states pivoting to a more pro-innovation stance that will benefit consumers.
While many states may press on and pass their versions of AI fairness laws alongside Colorado, Virginia’s and Texas’s course corrections indicate that the budding patchwork may not be simply a patchwork of compliance frameworks—but a competition between heavy-handed paperwork regimes versus pro-innovation, light-touch regulation frameworks.
Unfortunately, having a few safe havens does not guarantee AI can be sheltered from overbearing frameworks by locating in friendly states. Many AI Fairness laws define “developers” as any entity doing business within their state. Therefore, any AI firm seeking to do business in states with these laws would still have to comply with heavy-handed regulation—regardless of where it is headquartered. While this regulatory landscape can still harm innovation, a patchwork where some states get regulation right leaves the door open for at least some of AI’s benefits to reach consumers. Virginia and Texas demonstrate that states can and should think twice before passing unfinished, harmful regulations.
- States Should Not Wait for Federal Preemption
After signing Colorado’s AI Fairness bill, Governor Jared Polis called for federal action to preempt Colorado’s new law before it can do damage. However, the federal government’s AI policy transformed dramatically since 2024 when Governor Polis called for preemption. With the sudden about-face towards innovation and away from regulation, federal lawmaking reset, and DC has yet to demonstrate leadership in AI policy.
States already pursuing pro-innovation policies remain uninterrupted. Utah and Montana, for example, started from a pro-consumer standpoint and became leaders in AI policy—regardless of the federal government’s stance. As states begin to steer away from heavy-handed European-style regulation and the federal government begins crafting new AI policy, states need not wait to correct their algorithmic discrimination bills.
Conclusion
Virginia Governor Youngkin’s veto is a step in the right direction for AI policy and instructive for future action. Policymakers should consider how burdensome and redundant regulations affect consumers, understand that the AI industry moves faster than any lawmaking, and that each state can avoid a patchwork of heavy-handed regulatory regimes. Getting legislation right the first time around, rather than passing harmful laws for the sake of swift action, gives AI the space it needs to grow, promotes competition between big and small firms, and benefits consumers through better and more diverse AI offerings.
Nate Karren is a policy analyst with the American Consumer Institute, a nonprofit education and research organization. For more information about the Institute, follow us on X @ConsumerPal.