The three bars in the DMV — Virginia, DC, and Maryland — all operate under variations of the ABA Model Rules of Professional Conduct. All three have issued guidance on lawyer use of AI. The rules converge on a small set of obligations that, once understood, are very compatible with responsible AI deployment in a law practice.
The four rules that matter most
- Competence (Rule 1.1). Lawyers must provide competent representation, which includes keeping current with technology. AI use is now part of the competence landscape.
- Confidentiality (Rule 1.6). Lawyers must protect client confidential information. AI vendors that touch confidential material must be bound to appropriate confidentiality obligations.
- Supervision (Rules 5.1 / 5.3). Lawyers must supervise non-lawyer assistants, including AI tools. Supervision means knowing what the tool does, reviewing its output, and accepting responsibility for final work product.
- Fees and expenses (Rule 1.5). Lawyers cannot charge unreasonable fees. AI-assisted work that took significantly less time should be billed honestly.
How those rules translate architecturally
- For confidentiality: private or access-controlled deployment, no training on firm data, BAA-equivalent contracts with vendors, minimum-necessary data flow. See Attorney-client privilege and AI.
- For supervision: review checkpoints before AI-generated content reaches clients, courts, or opposing parties. Attorneys stay responsible for final work product.
- For competence: understanding what the AI does and doesn't do. Not every attorney needs to build AI — but every attorney using it needs to understand its scope and limits.
- For fees: transparent billing when AI has compressed work. If a matter took two hours instead of eight, clients are entitled to honest billing.
Jurisdictional variations
Virginia, DC, and Maryland rules are substantively similar on these points, with minor wording differences. Formal ethics opinions have been issued in each jurisdiction on topics like confidentiality with cloud vendors and the disclosure of AI use. The prudent posture is to satisfy the strictest applicable rule across all three jurisdictions.
Disclosure to clients
Whether to disclose AI use to clients is a fact-specific question. For most intake, drafting assistance, and administrative workflows, explicit disclosure isn't required as long as attorney review is in place. For more substantive uses, firms should consider disclosure as a matter of client trust even when not strictly required.
This post is not legal advice. Your firm's ethics counsel makes the final call. What we can do is build deployments designed to fit cleanly inside the rules. Scope an engagement.