ALEC Action Opposition to Virginia HB2094 (2025)

February 28, 2025

The Honorable Glenn Youngkin
1111 East Broad Street, 3rd Floor
Richmond, Virginia 23219

Re: Opposition to Virginia HB2094 (2025) – High-risk artificial intelligence; development, deployment, and use, civil penalties.

Dear Governor Youngkin,

On behalf of ALEC Action, I am writing today to express our opposition to Virginia HB2094, the High-Risk Artificial Intelligence Developer and Deployer Act. We encourage you to continue your Administration’s consistent and principled regulatory approach, pursuing policies that maximize the benefits of artificial intelligence (AI) and emerging technologies for all Virginians, while relying on existing state consumer protection and anti-discrimination laws to combat proven and specific harms.

Similar to legislation enacted in Colorado and rejected in Connecticut, HB2094 would create a “reasonable duty of care” standard for both AI developers and deployers to protect consumers from any known or foreseeable bias when AI is a substantial factor in making a so-called “consequential decision” across nine protected categories. While the patrons may be well- intentioned in their e orts to address the potential risks of novel AI technology, HB2094’s strict regulatory scheme targeting individuals and businesses in the name of preventing algorithmic bias would significantly undermine the Commonwealth’s position as a national leader in AI. When regulating emerging technology, legislation should be narrowly tailored and focused on harmful conduct, not the underlying technology or AI models themselves. State laws regulating AI tools should also use precise definitions that do not “inadvertently stymie or prohibit innocuous instances of machine learning.” Unfortunately, HB2094’s overly broad definitions of critical terms—such as “artificial intelligence,” “consequential decision,” “intentional and substantial modification,” and “substantial factor”—threaten to cover nearly all modern computing software and applications that use machine learning or algorithms in some way.

If enacted, HB2094 would be particularly problematic for startups, small businesses, and medium- sized businesses seeking to develop or integrate AI tools in the marketplace. The venture capital firm Andreesen Horowitz notes that policies like HB2094 punish what they call “Little Tech,” and insulate “Big Tech” incumbents—which have extensive resources to comply with a burdensome regulatory regime—from market competition. Virginia should not be in the business of picking winners and losers by burying new market entrants in unnecessary red tape and compliance paperwork.

Members of ALEC’s Communications and Technology Task Force developed an alternative regulatory approach that would ease regulatory burdens, promote competition in the AI marketplace, and build on the Commonwealth’s early successes as a hub for tech innovation.

Inspired by an AI framework implemented in Utah last year, the ALEC Model State Artificial Intelligence Act establishes a groundbreaking Artificial Intelligence Learning Laboratory program tasked with studying public policies that encourage AI development, identifying regulatory barriers impeding innovation, and evaluating the e ectiveness and viability of proposed regulation before the legislature.

The AI Learning Lab also conducts research, convenes stakeholders across multiple disciplines and perspectives, and provides for regulatory mitigation agreements that make it easier for startups to responsibly test cutting-edge innovations in a controlled environment, paving the way for broader regulatory reform if the tests are successful. Utah’s law also added reasonable transparency provisions that importantly make individuals responsible for their conduct, not the original developers of a large language model.

By incorporating some of these lighter-touch elements pioneered in Utah, you can avoid the unintended consequences of algorithmic bias legislation and solidify the Commonwealth’s legacy as a global leader in the responsible implementation of emerging technology. Under your leadership, Virginia has become a top state for AI development by implementing policies that build state AI expertise at agencies and within the General Assembly, encourage the responsible integration of AI tools across agencies, and protect consumers from specific threats. This includes:

• Executive Order 30, which established AI Policy Standards to streamline the AI procurement process across state government, and established guidelines for AI in education that embrace technological innovation while prioritizing student privacy and mitigating discriminatory outcomes.
• Executive Order 46, which banned the use of the foreign adversary-owned AI model DeepSeek on government-owned or government-leased devices, or on government networks.
• Virginia HB2678 (2019) and SB731 (2024), which were among the first state laws in the nation to combat deepfake child sexual abuse material (CSAM) and the non-consensual distribution of intimate deepfake media.

Recent breakthroughs in AI and machine learning are already ushering in a new era of digital transformation in sectors like health care, education, financial services, human resources, legal services, and software development – many of the same “consequential decision” categories that HB2094 seeks to heavily regulate. It would be a mistake for the Commonwealth to suddenly reverse its longstanding position and overregulate AI technology at this critical moment.

For these reasons, we respectfully urge you to veto HB2094 in its current form.

Sincerely,

Andrew Yates
President
ALEC Action