AI Law Framework

Jurisdiction

Vermont: AI rules for lawyers

Vermont, United States · Switch state

What this means for lawyers

In Vermont state courts you may use AI if the use and the work product meet legal and ethical standards. If you give AI output to the Judiciary, or use it in a proceeding, you are expected to review it and confirm that it is accurate. You must not let the tool compromise sensitive or confidential information.

The Vermont Rules of Professional Conduct apply in full to a lawyer's use of AI systems and output. Administrative Directive No. 38 does not require you to tell the court that you used AI.

The judiciary committee's first annual report, dated 3-1-25, recommends no amendments to court rules, the Vermont Rules of Professional Conduct, or the Code of Judicial Conduct. It treats existing rules as covering use and misuse of generative AI, and it summarises guidance rather than issuing a formal ethics opinion. Competence includes knowing the risks and benefits of generative AI.

An unchecked quotation can breach Rule 11. In LaRoche v. Sterett (LaRoche), No. 24-AP-406, the Supreme Court held that counsel violated V.R.C.P. 11(b)(2) by filing a quotation, attributed to a real case, that counsel said came from a client's "AI helper" draft and that counsel did not check. The sanction is to file the entry order in pending Superior Court cases and to certify that within 14 days.

As at .

Telling the court and the client

  • Tell the courtUnclear

    Court disclosure: no rule found. Administrative Directive No. 38 requires court users to review and confirm the accuracy of AI output given to the Judiciary or used in a proceeding. It does not require disclosure or a certification that AI was used.

    The 2025 committee recommended no court-rule amendment on disclosure, transparency, or certification. Client disclosure/consent: depends. The committee's summarised guidance says a lawyer might consider disclosing to the client the intent, nature, and extent of generative-AI use, and it urges caution if a lawyer intends to rely solely on generative AI to communicate with a client.

    That is a recommendation in a committee report. The report does not state that client consent is required before any use. The same summary says use can risk an unauthorised disclosure of confidential information and includes review of a tool's terms of service.

    Administrative Directive No. 38, Policy on Artificial Intelligence
    First Annual Report
  • Tell the clientVaries

    Same basis and sources as above.

Choose a practice areaAll US federal courts, states and DCEvery jurisdiction

What changed in Vermont in the last 12 months

No new or changed AI rules recorded for Vermont in the last 12 months.

AI instruments

Full notes

Courts and tribunals

First Annual Report

3-1-25; start date not confirmedProfessional guidanceVermont Judiciary Committee on Artificial Intelligence and the Courts3-1-25Last checked 31 October 2026

Sources

16 sources, each with its issuer and the date it was last checked.Show all sources

Before the work starts

  • Define the matter

    Confirm the client’s instructions, jurisdiction, intended output and the information available before dividing the work.

  • Confirm the source material

    Check the work against the original records and identify omissions, contradictions and remaining information.

  • Record review and approval

    Resolve exceptions and record the responsible lawyer’s approval of the exact output and intended use.

  • Jurisdiction

    Apply this guidance to the actual matter and jurisdiction. Confirm current requirements before relying on an example.

Limits on AI assistance

  • Confirm firm, client and applicable court or professional restrictions before using any AI feature.
United StatesUnited States federal courtsAlabamaAlaskaArizonaArkansasCaliforniaColoradoConnecticutDelawareDistrict of ColumbiaFlorida