AI Anti-Discrimination Bill Heads to Governor’s Desk in Virginia

On February 20, the Virginia State Senate passed their own AI anti-discrimination law, House Bill 2094, that if signed by Governor Glenn Youngkin, would position Virginia to join Colorado in enacting AI anti-discrimination legislation.

Compliance Obligations Under the Virginia Bill

Virginia’s “High-Risk Artificial Intelligence Developer and Deployer Act” (Act) largely resembles the efforts of Colorado’s AI law. If signed, the Act would take effect on July 1, 2026, and introduce new legal responsibilities for “deployers” and “developers” of “high-risk” AI systems, focusing on mitigating algorithmic discrimination where high-risk AI systems are used to make consequential decisions. “High risk” AI systems are defined as those intended to “autonomously make, or be a substantial factor in making, a consequential decision.” This includes, for example, situations where AI is used to make decisions relating to an individual’s access to employment, insurance, financial or lending services, to name a few.

Key provisions include:

• Duty of Reasonable Care: Both developers and deployers of high-risk AI systems must exercise a duty of reasonable care to protect consumers from algorithmic discrimination-related risks, whether known or reasonably foreseeable.
• Disclosures: Developers must provide deployers with documentation detailing mitigation efforts and potential discrimination risks associated with these systems.
• Risk Management Policies: Deployers are required to establish and implement a risk management policy designed to “identify, mitigate, and document” algorithmic discrimination in consequential decision-making of these systems.
• Impact Assessments: Before deploying a high-risk AI system or making significant updates to a system, deployers must conduct impact assessments to ensure transparency about the system’s benefits and risks, including algorithmic discrimination. Deployers must also disclose performance metrics and inform consumers when they are engaging with an AI system, providing the consumer with relevant system information. Further, if the deployer uses the system to make consequential decisions about consumers, the deployer is required to inform the consumer. If the decision is adverse and based on external personal data, deployers are required to provide the reason for the decision, opportunities to correct inaccuracies in data, and offer an appeal process that allows for human review.

Enforcement

The Act would empower the state Attorney General to enforce the Act, and also launch civil investigations where there is reasonable cause of a violation. Notably, the Act does not create a private cause of action for individuals, even where they have been impacted by a consequential decision of a high-risk AI system.

Where the violation is found to be willful, civil penalties for each separate violation may be as high as $10,000. Other violations carry penalties of up to $1,000. However, the Act establishes certain affirmative defenses are available to developers and deployers.

Key Takeaways

With Virginia’s legislature passing this comprehensive AI anti-discrimination law, it awaits Governor Youngkin’s signature to become law. The Governor has not indicated whether he intends to sign the bill but has until March 24 to act on the legislation. If enacted, the law will impose new compliance obligations on deployers and developers of high-risk AI systems in Virginia, including the requirement to perform comprehensive impact assessments. Pending a signature from the Governor, deployers and developers should familiarize themselves with these new obligations as well as the affirmative defenses available to them before the Act’s effective date on July 1, 2026.

The AI, Metaverse, & Blockchain industry group at ArentFox Schiff will be closely monitoring legislative developments with respect to Virginia’s AI anti-discrimination bill. If you have any questions or concerns pertaining to the Act or its potential implications on your business, please reach out to members of the ArentFox Schiff AI, Metaverse, & Blockchain team.