October 18, 2026 (in 9 days): Tenth Circuit: comments close on 2027 rules, incl. proposed Rule 46.5(B)(4) generative-AI review →

D. Colo.: AI Guidance (Hon. Maritza Dominguez Braswell, D. Colo.)

Hon. Maritza Dominguez Braswell, U.S. Magistrate Judge · U.S. District Court, District of Colorado

active

Verified September 30, 2026

Citation
AI Guidance (Hon. Maritza Dominguez Braswell, D. Colo.)
Order date
December 1, 2025

Status and scope

No AI-specific filing obligation: in force, but it imposes no disclosure, verification, or use restriction on attorney filings beyond existing rules. It provides: Document is styled as 'guidance,' not a strict standing order, and does not impose a new pre-filing certification requirement.

No AI-specific filing obligation. This instrument is in force but imposes no disclosure, verification, or use restriction on attorney filings beyond existing rules; where it restates duties that already apply (Rule 11, professional conduct rules), those duties remain binding. It is listed so firms can see the court's stated expectations; the items below are not AI-specific filing obligations.

What the instrument provides

Practice areas: federal civil

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 rule requires

Magistrate Judge Braswell’s AI Guidance occupies a different posture from Judges Wang, Crews, and Prose’s standing orders in the same district. The document is explicitly styled as guidance rather than a procedural rule, and it does not impose a new pre-filing certification requirement. The premise is that “the source of error may change with technology, but the same rules apply”: existing duties of cite checking, verification, accuracy, and candor already cover hallucinations as a category of inaccurate statement.

The guidance’s operative content is a procedural framework for the situation that arises when an attorney encounters a hallucination in an opponent’s filing. The framework is three steps: confer with the Drafter to confirm the error and explore correction; inform the Court via joint email to BraswellChambers@cod.uscourts.gov; escalate by formal Notice if the Drafter is unresponsive or refuses to engage in good faith.

When the Court’s involvement is required, the guidance lists nine non-exhaustive factors the Court will weigh: whether the Drafter promptly engaged in good faith; took responsibility; took immediate steps to mitigate harm; follows an AI Use Policy demonstrating responsible practices; took reasonable steps to verify accuracy before filing; whether the Drafter is pro se (impacting the reasonable-steps assessment); the nature and extent of the hallucinations (which may indicate recklessness); whether the hallucination caused meaningful prejudice or harm to anyone; and whether the hallucination resulted in a significant waste of judicial resources. The guidance’s closing remarks expressly state: “This document is not an order and does not establish new rules or requirements.”

The guidance fills a notable gap in district practice: it tells lawyers what to do when their adversary files a hallucinated brief, rather than what to do before filing their own. Counsel handling matters before Judge Braswell should pre-load the conferral email address into chamber-specific protocols.

R&G data corrections

R&G’s tracker dates this entry 2025-12-01. The document bears no date on its face. Its embedded PDF creation date is October 29, 2025, and it first appears on Judge Braswell’s chambers page in a Wayback Machine capture of December 9, 2025 (it is absent from the July 20, 2025 capture). The 2025-12-01 order_date is the Ropes & Gray tracker’s date.

Quotable language

“With this document, the Court provides guidance to the parties practicing before it. The guidance is focused on the issue of ‘hallucinations,’ but also seeks to move the conversation beyond that.”

“The source of error may change with technology, but the same rules apply. Cite checking is still critical. Verification is still necessary. Accuracy, candor, integrity, and competence are all still cornerstones of our professional responsibility.”

“There is an important distinction between the reckless use of AI, and ordinary human error in the face of change. This is not a call to set aside accountability. It is a reminder that learning curves are real, and a culture that balances accountability with grace will strengthen our profession and better serve the public we are all here to help.”

Primary source

AI Guidance (Braswell, M.J.), cod.uscourts.gov