HB1294 would amend Virginia law governing the use of artificial intelligence in criminal justice and law-enforcement investigations. The bill defines “artificial intelligence-based tool” broadly and expands the concept of “covered artificial intelligence” to include a wide range of investigative technologies, such as biometric identification, forensic DNA phenotyping, investigative genetic genealogy, automated license plate readers, gunshot detection, predictive policing, risk scoring, sentiment analysis, fraud detection, financial transaction analysis, and social media analysis. It also distinguishes these tools from ordinary administrative software and basic grammar or spelling assistance.
The bill requires that major criminal-justice decisions—such as pretrial detention, release, prosecution, sentencing, probation, parole, correctional supervision, and rehabilitation—remain under human control, and it allows any AI-generated recommendation or prediction to be challenged as permitted by law. It further requires disclosure when covered AI is used in a criminal investigation, including in police reports and to prosecutors and defense counsel, and imposes additional requirements for reports or records created with generative AI, such as disclaimers, identification of AI-generated content where feasible, author certification, retention of drafts, and audit trails.
HB1294 would also create enforcement mechanisms. The Attorney General could investigate and seek equitable or declaratory relief against a law-enforcement agency for violations, and private individuals within the agency’s jurisdiction could sue for enforcement after giving 90 days’ written notice and an opportunity to cure. This would add new transparency, recordkeeping, and accountability obligations for law-enforcement agencies using AI in investigations and report writing.
The bill appears to reflect a cautious or skeptical approach to AI in policing, emphasizing human oversight, disclosure, and the ability to challenge AI-derived outputs. Its last recorded action was that it was continued to the next session in the Communications, Technology and Innovation Committee by voice vote, suggesting the measure did not advance immediately and may have faced unresolved policy or implementation concerns.
No committee transcript or recorded votes were provided, so specific points of contention are not documented in the available materials. Based on the bill’s structure, likely areas of debate include the breadth of the disclosure requirements, the administrative burden on law-enforcement agencies, the scope of technologies covered, and the private right of action with attorney-fee recovery.
HB1294 would amend § 19.2-11.14 of the Code of Virginia to impose new disclosure, documentation, and human-oversight requirements on law-enforcement use of artificial intelligence in criminal investigations and related records. It would affect police departments, prosecutors, defense counsel, courts, and individuals under investigation by requiring notice of AI use, retention of drafts and audit trails, and limitations on fully automated decision-making in criminal justice outcomes.
The available record suggests a generally cautious and oversight-oriented sentiment toward the bill, with the proposal framed around transparency, accountability, and human decision-making in criminal justice. However, the fact that it was continued to the next session in committee indicates that support was not sufficient for immediate advancement, or that members wanted more time to evaluate the policy and operational implications.
No direct transcript evidence identifies specific disagreements, but the bill’s likely points of contention are the breadth of the definition of covered AI, the requirement to disclose AI use in police reports and to defense counsel, the retention and audit-trail mandates for generative AI, and the enforcement provisions allowing both Attorney General action and private lawsuits. Law-enforcement agencies may view these provisions as burdensome or difficult to implement, while supporters are likely to argue they are necessary to prevent opaque or unreviewable AI use in criminal investigations.