Introduction
Virginia is taking a proactive approach in regulating artificial intelligence (AI) through the introduction of two significant bills by Delegate Michelle Lopes Maldonado. These bills [3], which have been approved by the General Assembly and are awaiting the governor’s signature [2], aim to address the regulation of high-risk AI systems and the legal implications of synthetic digital content.
Description
Virginia Delegate Michelle Lopes Maldonado has introduced two significant AI bills that have been approved by the General Assembly and are awaiting the governor’s signature [2]. These proposed regulations reflect Virginia’s proactive stance in technology and privacy law, aiming to regulate high-risk AI systems and their developers, integrators [1], and deployers [1] [3].
The first bill [2], HB 2094 [2] [3], seeks to impose comprehensive regulations on high-risk AI systems. It mandates that developers provide detailed information regarding the system’s intended use, limitations [1] [2], risks of algorithmic discrimination [2] [3], performance evaluations [2], and risk mitigation strategies [2]. Furthermore, developers are required to conduct impact assessments and implement risk management policies [3]. The bill also includes transparency requirements for synthetic content generated by high-risk generative AI applications [2], a discretionary right to cure [2], and stipulates that the state attorney general must issue a civil investigative demand before any enforcement actions can be initiated [2]. An economic analysis by Chamber of Progress estimates that compliance with this legislation could cost Virginia’s AI innovators approximately $290 million [3], based on data from the European Union AI Act [3], which indicates that AI developers typically allocate about 17.22 percent of total model development costs to compliance [3]. The bill has successfully passed the House and is currently under review by the Senate Committee on General Laws and Technology [3]. Concerns have been raised regarding the potential impact of such regulations on Virginia’s status as an AI hub [3], drawing comparisons to similar legislation in states like Colorado, which is now reconsidering some of its requirements during the implementation phase [3].
The second bill [2], HB 2124 [2], known as the Synthetic Digital Content Act [2], extends existing defamation laws to encompass synthetic digital content [2]. It establishes a Class 1 misdemeanor for using AI-generated content to commit fraud and allows individuals depicted in such content to pursue civil actions against violators [2]. Additionally, the bill directs the Attorney General to form a work group to evaluate the enforcement of laws concerning synthetic digital content [2].
Looking ahead [2], AI policy leaders are preparing for future legislative opportunities [2], as two other AI-related bills introduced by Delegate Maldonado have been referred to the Joint Commission on Technology and Science for further study [2]. These include HB 2121 [2], which requires AI developers to apply provenance data to AI-generated digital content [2], and HB 2250 [2], which allows consumers to opt out of having their personal data used for AI training [2], while also mandating mechanisms for data verification and deletion requests [2]. With a shorter legislative session approaching in 2025 [1], the outcome of these AI bills is expected to be determined soon [1].
Conclusion
The introduction of these AI bills signifies Virginia’s commitment to leading in the realm of technology and privacy law. By addressing the regulation of high-risk AI systems and the legal implications of synthetic digital content, Virginia is setting a precedent for other states. The potential economic impact and the state’s status as an AI hub are key considerations as these regulations move forward. As the legislative process continues, the outcomes of these bills will likely influence future AI policy and innovation both within Virginia and beyond.
References
[1] https://www.woodsrogers.com/insights/publications/virginia-ai-bills-could-serve-as-nationwide-model
[2] https://www.transparencycoalition.ai/news/victory-in-virginia-the-old-dominion-approves-two-important-ai-bills
[3] https://progresschamber.org/new-analysis-ai-discrimination-bill-may-cost-virginia-developers-and-deployers-millions/
