A BILL to amend the Code of Virginia by adding a section numbered 54.1-111.1, relating to professions and occupations; impersonation of certain licensed professionals by chatbot; notice; civil liability.
HB669 creates a new section of the Virginia Code aimed at preventing artificial intelligence systems from impersonating licensed professionals or otherwise providing services that would require a state license if performed by a person. The bill defines “artificial intelligence system,” “developer,” and “deployer,” and then prohibits deployers from making publicly available AI systems that falsely present themselves as licensed professionals, provide individualized diagnosis/treatment/advice/services requiring licensure, or otherwise engage in unlicensed practice in fields covered by a long list of Virginia licensing statutes.
The bill also requires clear notice when a reasonable user would think they are interacting with a human rather than an AI system. That notice must be conspicuous, explicit, in the same language as the system, and displayed in a readable size. In most cases, providing the notice is a defense to enforcement, but it does not protect systems that are actually impersonating licensed professionals or delivering individualized professional services that require licensure. Enforcement is assigned exclusively to the Attorney General, who must give 30 days’ written notice and an opportunity to cure before filing suit. If violations continue, the Attorney General may seek injunctions and civil penalties of up to $7,500 per violation, with recovered funds directed to a state enforcement trust fund.
In practical terms, the bill would add a new layer of consumer protection and professional-regulation oversight to Virginia’s AI marketplace. It would affect AI developers and deployers operating in the Commonwealth, especially chatbot and other interactive systems that present themselves as human or that offer advice in regulated professions such as health care, counseling, education-related services, and other licensed occupations. The bill expressly excludes certain ordinary software tools and internal business functions from the definition of AI system.
The general sentiment reflected by the bill’s structure is precautionary and consumer-protective, with an emphasis on preventing deception and unlicensed practice rather than broadly restricting AI use. Because there were no recorded committee transcripts or votes provided, there is no direct evidence of debate or bipartisan support/opposition in the materials supplied. The bill’s placement in Appropriations and its status of being left in committee suggest it did not advance, but the available record does not identify the specific reason.
The main point of contention likely concerns how far the state should go in regulating AI-generated professional advice and whether the notice requirement and enforcement framework are sufficient or too restrictive. Potential concerns include the breadth of the definition of AI system, the scope of professions covered, compliance burdens on businesses, and whether the bill could chill legitimate AI customer-service or informational tools. Supporters would likely emphasize fraud prevention, public safety, and protecting consumers from being misled into relying on AI as if it were a licensed professional.
The bill would add § 54.1-111.1 to the Code of Virginia and create new restrictions on AI systems that impersonate licensed professionals or provide individualized services reserved to licensed practitioners. It would also establish a mandatory disclosure regime for AI systems that could be mistaken for humans, authorize exclusive enforcement by the Attorney General, create a 30-day cure process, and impose civil penalties and fee recovery for violations. The measure would primarily affect AI deployers, developers doing business in Virginia, and businesses offering chatbot or other interactive AI services in regulated professional contexts.
The bill appears to have been framed in a consumer-protection and professional-licensing context, suggesting a generally cautious or skeptical posture toward AI systems that could mislead users. No committee testimony or vote record was provided, so there is no direct evidence of support or opposition from the legislative record included here. Its failure to move beyond Appropriations indicates it did not gain enough traction to advance, but the materials do not explain whether that was due to policy disagreement, fiscal concerns, or procedural reasons.
The likely areas of contention are the scope of the prohibition and the compliance burden on AI businesses. Opponents could argue that the bill’s definitions are broad, that the notice requirements may be difficult to implement across platforms and languages, and that the restrictions could sweep in legitimate informational tools that are not actually practicing a profession. Supporters would likely focus on the risk of consumers being deceived by AI systems into believing they are receiving advice from licensed professionals, especially in health, counseling, and other regulated fields. The bill’s carveout for ordinary software and its exclusive enforcement by the Attorney General suggest an attempt to narrow the reach and avoid private litigation.