District of Columbia: AI rules for lawyers

What this means for lawyers

D.C. Bar Ethics Opinion 388 (published April 2024) is the main D.C. guidance: competence means understanding enough about generative AI to be reasonably confident it will help the client, and AI output is not a substitute for ordinary fact- and cite-checking.

Do not put client confidences or secrets into a generative AI tool unless satisfied (by privacy terms or negotiated terms) that the provider will not see them or reveal them to later users; the opinion warns that anonymising may not be enough and that clients typically should not be asked to consent to wide-ranging disclosure.

Candor and fees: fake or false citations engage Rule 3.3; breaching a tribunal's AI order engages Rule 3.4(c); hourly-fee lawyers may bill only time actually spent, and AI costs can be passed through only as agreed expenses.

The D.C. Court of Appeals amended the Rules of Professional Conduct by order filed 7 April 2025, effective 6 May 2025. The amendments include Rule 1.1 Comment [5] (duty to stay abreast of technological changes), Rule 1.6 (protecting client information), and the Rule 1.1 and 5.3 comments on outsourcing. Read the amended Comment [5] wording before relying on it.

As at 31 October 2026.

Telling the court and the client

• Tell the courtUnclear

  Court disclosure: no rule found (general D.C. rule). Opinion 388 refers to tribunals that require disclosure of AI use but names no D.C. rule imposing it; individual judge orders could not be checked. Client disclosure/consent: depends.

  Opinion 388 requires informed client consent before confidential information is shared with an AI provider that would expose it, and says an expected out-of-pocket AI cost that will be billed should be communicated under Rule 1.5(b). It does not state a general duty to tell clients about AI use.

  Ethics Opinion 388
• Tell the clientVaries

  Same basis and sources as above.