A BILL to amend the Code of Virginia by adding in Article 1 of Chapter 1 of Title 9.1 a section numbered 9.1-116.11, relating to law enforcement; artificial intelligence inventory; civil action.
HB1295 would require every state and local law-enforcement agency and sheriff’s department in Virginia to conduct an annual inventory of any “covered AI system” it uses and to publish that inventory by November 1 each year. The bill defines covered AI systems broadly to include tools used in investigations or report writing, such as biometric identification, forensic DNA phenotyping, investigative genetic genealogy, cross-camera tracking, automated license plate readers, gunshot detection, predictive policing, risk scoring, behavioral analysis, fraud detection, social media analysis, and other AI-enabled analytics platforms.
For each covered system, agencies would have to disclose the system’s name, vendor and product name if applicable, a brief description of its capabilities and limitations, the types of data inputs and outputs, and the authorized and unauthorized uses. The bill excludes general administrative tools and limited generative AI uses like spelling or grammar checks, as well as AI used only for development, prototyping, or research before deployment. It also creates enforcement mechanisms allowing the Attorney General to seek equitable or declaratory relief and allowing residents within an agency’s jurisdiction to sue after giving 90 days’ written notice, with prevailing plaintiffs eligible for attorney fees and costs.
The bill would add a new section, § 9.1-116.11, to Title 9.1 of the Code of Virginia and impose a new transparency and reporting obligation on law-enforcement agencies and sheriff’s departments that use AI in investigative or operational contexts. It would not ban AI use, but it would require public disclosure of AI inventories and system details, potentially affecting procurement, documentation, and internal governance practices for agencies using surveillance, analytics, or investigative technologies. The bill also expands enforcement exposure by authorizing civil actions by both the Attorney General and private residents.
The available context suggests the bill was treated cautiously and did not advance, as it was continued to the next session in the Communications, Technology and Innovation Committee by voice vote. With no recorded floor votes or committee transcript excerpts provided, there is no direct evidence of strong public debate in the record here, but the committee outcome indicates at least some hesitation or need for further review. Overall, the bill appears to have been framed as a transparency measure rather than a prohibition, which may have made it more palatable to some members while still raising concerns for others.
The main points of contention are likely to be the breadth of the AI definition, the scope of systems covered, and the compliance burden on law-enforcement agencies. Agencies may object to having to publicly disclose investigative tools, capabilities, and limitations, especially for technologies tied to surveillance, predictive analytics, or criminal investigations. Privacy advocates and transparency supporters would likely favor the bill’s disclosure requirements and private right of action, while law-enforcement stakeholders may be concerned about operational security, vendor confidentiality, and litigation risk from the Attorney General or private plaintiffs.