October 18, 2026 (in 16 days): Tenth Circuit: comments close on 2027 rules, incl. proposed Rule 46.5(B)(4) generative-AI review →

Idaho OAH: Policy 25-1: Guidelines for Administrative Law Judges Regarding the Use of Art…

Issued by the OAH's chief, Bryan Nickels · State of Idaho, Office of Administrative Hearings

active

Verified September 28, 2026

Citation
Policy 25-1: Guidelines for Administrative Law Judges Regarding the Use of Artificial Intelligence (Idaho Office of Administrative Hearings)
Order date
September 24, 2025

Status and scope

Applies to the issuing body's own personnel or adjudicators, not to attorney filings, and imposes no obligation on attorneys. It provides: Applies to OAH Administrative Law Judges, independent contract hearing officers, and 'anyone in roles supporting OAH's ALJs and ICHOs, including deputy clerks and legal assistants.' It imposes no duty on parties or counsel.

Does not apply to attorney filings. This instrument binds the issuing body's own personnel or adjudicators. It is tracked for completeness and imposes no obligation on attorneys.

What the instrument requires of its addressees

Practice areas: state administrative

Verify this order against the court's official website before relying on it. Standing orders are amended without notice. Requirements vary by judge and case type.

What the guidelines do

Idaho’s Office of Administrative Hearings issued Policy 25-1 on September 24, 2025: “Guidelines for Administrative Law Judges Regarding the Use of Artificial Intelligence.” The OAH’s announcement states that the guidelines replace “existing OAH Policies 23-1 and 23-12,” the 2023 interim ban on AI in drafting orders and a 2023 policy on AI disclosure statements from parties.

The guidelines govern the OAH’s adjudicators, not the parties before them. They apply to Administrative Law Judges, to independent contract hearing officers, and to “anyone in roles supporting OAH’s ALJs and ICHOs, including deputy clerks and legal assistants.” They adopt the New Jersey courts’ AI glossary for terminology and describe themselves as “not intended to be final.”

Prohibited uses in case proceedings

Three core rules bar AI from adjudicative work:

  1. Decisions. “ALJs are not permitted to use AI/GenAI to write orders or decisions.” The ban reaches “any stage of drafting orders and/or decisions, including the drafting of outlines or similar pre-final documents,” and draws on the state Office of Information Technology Services’ prohibition on AI for “Fully autonomous decision-making affecting individual rights, benefits, or services.”
  2. Legal research. “ALJs are not permitted to use AI/GenAI for legal research in case proceedings,” whether through public AI tools or AI features of legal-research vendors, because of the risk of “hallucinated case citations, mis-quotes, or inaccurate/incomplete summaries.”
  3. Evidence and submissions. “ALJs are not permitted to use AI/GenAI for analysis of evidence or party submissions in case proceedings.” Every party submission “must be reviewed by the ALJ independent of any computerized summary,” and the guidelines cite the privacy risk of uploading personally identifiable information.

The OAH as an agency also will not use AI in functions tied to individual cases, such as chatbots, scheduling, or translation of testimony, submissions, and evidence.

Transcripts

ALJs may rely on AI-generated voice-to-text transcripts, as IDAPA 62.01.01.601.04 already allows. Where a recording system lacks automatic transcription, a transcript may be made with State of Idaho-licensed software that retains neither the audio nor the transcript; ALJs may not use unlicensed or public AI systems to generate transcripts. Such a transcript is unofficial: ALJs must cite the hour and minute of the recording, verify each citation against it, and not use AI-generated hearing summaries.

What parties and counsel should know

The guidelines impose no duty on litigants directly, but Guideline 5 preserves an ALJ’s power to require one. ALJs “are permitted to request disclosure statements from parties (whether represented or self-represented) regarding the use of AI/GenAI in the preparation of briefing and other written submissions, including evidence,” at any time and “solely within the discretion of the ALJ.” The recommended party certification reads:

“I certify that no portion of this filing has been drafted by generative artificial intelligence, or otherwise that any language drafted by generative artificial intelligence contained herein—including quotations, citations, paraphrased assertions, and legal analysis—has been checked for accuracy, using print reporters or traditional legal databases, by a human being before it has been submitted to the Hearing Officer. I understand that by signing this filing, I am responsible for the contents herein, regardless of whether generative artificial intelligence drafted any portion of this filing.”

An alternative provision may be placed in an order before submissions are made. The guidelines also say ALJs “are not to use AI/GenAI or other artificial intelligence system to identify the use of AI/GenAI in a party’s submissions.”

Primary sources

How this order changed

Dated record of this order's own lifecycle. Where it was amended or withdrawn, the badge shows its standing after that step.

  1. active
    Issued as Policy 25-1, replacing OAH Policies 23-1 (interim ban on AI in drafting orders) and 23-12 (discretionary AI disclosure statements from parties).