Virginia issued formal AI ethics guidance in mid-2024. The Virginia State Bar published Legal Ethics Opinion 1901 in July 2024, directly addressing generative AI use by Virginia attorneys. The opinion covers competence, confidentiality, billing, and disclosure — providing a comprehensive framework anchored in Virginia's existing Rules of Professional Conduct.
Virginia's position matters beyond state lines. Multiple federal courts in Virginia's Eastern and Western Districts have adopted AI disclosure requirements for filings, and the bar's opinion aligns with these judicial expectations. For firms practicing in Virginia federal courts, the bar guidance and court rules create reinforcing obligations.
What the Bar Says
Legal Ethics Opinion 1901 (July 2024) maps AI obligations to Virginia's Rules of Professional Conduct. Rule 1.1 (Competence) requires lawyers to understand AI tools sufficiently to use them effectively and identify errors. The opinion explicitly addresses hallucination risk — attorneys must verify all AI-generated citations and legal analysis. Rule 5.3 (Supervision) applies to AI the same way it applies to non-lawyer assistants: the attorney must supervise, review, and take full responsibility. The opinion treats AI as a tool, not a delegate — the ethical obligations remain entirely with the lawyer.
Billing Implications
The opinion requires reasonable fees under Rule 1.5. AI tool costs may be passed through to clients as expenses, but only with client consent. This is a notable requirement — firms cannot simply add AI costs to invoices without disclosure. The opinion implies that billing at full attorney rates for work primarily completed by AI may not meet reasonableness standards. Firms should consider value-based billing for AI-assisted matters and maintain detailed records of which tasks involved AI assistance.
Confidentiality Rules
Virginia requires reasonable efforts to safeguard client information when using AI under Rule 1.6. The opinion mandates vendor evaluation — attorneys must understand how the AI provider handles, stores, and potentially shares data. Key questions include: Does the vendor use input data for model training? What are the data retention periods? Are there contractual confidentiality protections? The opinion recommends enterprise-grade tools over consumer AI and suggests that using free, consumer-facing AI tools with client data likely violates confidentiality obligations.
What's Still Unclear
The opinion does not define what constitutes "sufficient understanding" of AI tools under the competence requirement. It recommends disclosure but does not specify exactly when disclosure becomes mandatory versus aspirational. The interaction between LEO 1901 and various Virginia federal court AI disclosure orders is not addressed — attorneys must independently track court-specific requirements. Questions about malpractice coverage for AI-related errors and whether AI-specific CLE requirements will follow remain open.
The Bottom Line: Virginia's LEO 1901 provides clear, practical AI ethics guidance with a notable requirement for client consent before passing through AI tool costs.
AI-Assisted Research. This piece was researched and written with AI assistance, reviewed and edited by Manu Ayala. For deeper takes and the perspective behind the research, follow me on LinkedIn or email me directly.