What happened
Source New Mexico reports that Second Judicial District Family Court Judge Jane Levy and Sixth Judicial District Judge Jarod Hofacket told New Mexico lawmakers they are helping draft three sets of statewide guidelines for artificial intelligence use in court. The guidelines would address self-represented litigants, attorneys and judicial staff.
Source New Mexico reports that Judges Jane Levy and Jarod Hofacket briefed the interim Courts, Corrections and Justice Committee in Albuquerque on Tuesday about work underway on statewide guidance for artificial intelligence in court. The judges serve on the New Mexico Supreme Court’s Committee on Artificial Intelligence and the Courts, which also includes academics and state lawmakers.
According to the report, the committee is developing three distinct sets of guidelines: one for people representing themselves, one for attorneys and one for judicial staff. The source describes the effort as an attempt to clarify expectations for anyone interacting with the legal system. The judges’ central message, as quoted by Source New Mexico, was that AI should be treated as a tool rather than an authority. Levy compared it with spellcheck and framed the practical problem as ensuring that every lawyer appearing in the state understands the rules before each judge.
That distinction matters because the same system may be used for drafting, research or administrative work, while the responsibility for checking the result remains with the human participant. The report does not specify which AI systems are covered, whether the guidelines will distinguish between generative and non-generative tools, or how the proposed rules would treat confidential court information.
Source New Mexico cites two courtroom examples as reasons for guardrails. The report says an Arizona family used AI last year to recreate a deceased relative, who had died in a road-rage incident, so the relative could deliver courtroom statements. It also says the New Mexico Supreme Court recently found a local attorney in contempt of court for filing an AI-generated brief that contained fabricated witnesses and testimony. Those examples show different kinds of risk—synthetic representation in a courtroom and false legal material in a filing—but the source does not provide the names of the family or attorney, the exact sanctions beyond contempt, or technical details about the tools involved.
Read the primary source: sourcenm.com ↗
Why it matters
The effort could establish clearer expectations for people using AI in New Mexico’s legal system, where inaccurate or fabricated output can affect filings, testimony and public trust. The source reports that the New Mexico Supreme Court recently held a local attorney in contempt after an AI-assisted brief included fabricated witnesses and testimony.
Court proceedings depend on records, arguments and testimony that participants can examine and challenge. When an AI system supplies invented authorities, witnesses or testimony, the problem is not merely an ordinary drafting error: it can consume court and opposing counsel resources and make it harder to identify what evidence or law is actually before the court. Source New Mexico’s report connects the proposed guidelines directly to a New Mexico contempt finding involving fabricated material, but it does not independently confirm the underlying court record in the material provided here.
The proposed division into guidance for self-represented litigants, attorneys and judicial staff recognizes that people encounter AI from different positions of responsibility. A person navigating court without a lawyer may need plain-language warnings about checking AI-generated legal information. Lawyers may need rules about verification, filings and professional duties. Judicial staff may face separate questions about administrative use and handling court information.
These are practical implications of the categories reported by Source New Mexico; the article does not state the committee’s detailed recommendations or whether it has reached agreement on them. The issue also concerns the human role in judging. Hofacket told lawmakers that a human being listening to a person’s problems and deciding a case remains an essential part of the courtroom experience.
That statement is a view reported by Source New Mexico, not evidence that the committee has proposed replacing or preserving any particular judicial function through formal policy. Still, it identifies a broader public question: even when AI is used only for support, courts must make clear who is accountable for decisions and how participants can challenge inaccurate or misleading machine-generated material.
What to watch next
The judges’ committee is expected to produce annual recommendations, but the source does not say when the guidelines will be finalized, whether they will be binding, or what enforcement mechanisms will apply. It also does not independently establish how widespread AI use or AI-related courtroom errors are in New Mexico.
The next concrete milestone is the committee’s guidance. Source New Mexico reports that the Committee on Artificial Intelligence and the Courts is set to produce annual recommendations, but the article does not give a publication date for the first full guidelines. Watch for whether the recommendations are advisory or incorporated into court rules, practice standards or filing requirements. The distinction will determine how much practical force they have for litigants, lawyers and court employees.
The content of the guidance will also matter. Key unanswered questions include whether users must disclose AI assistance, how lawyers must verify citations and factual claims, whether courts will restrict AI-generated testimony or reconstructed statements, and what safeguards will apply to confidential or sensitive information. None of those details is provided in the source, so they should not be treated as established features of the New Mexico plan.
The source also does not say whether the committee has consulted affected litigants or tested draft rules in court. Finally, officials and the public will need evidence about how the policy works in practice. Relevant follow-up would include the final text of the recommendations, any New Mexico Supreme Court action, implementation dates, training for court personnel and lawyers, and information about complaints or sanctions involving AI use.
This report is based on Source New Mexico’s account of the judges’ presentation. No public primary document, final guideline or independent confirmation of the reported incidents is supplied here, so the scope and eventual effect of the initiative remain uncertain.
