N.D. Ill.: Case Management Procedures, AI Provision (Judge Sharon Johnson Coleman, N.D. I…

What the order requires

Judge Coleman’s case management procedures anchor AI use in the Federal Rules rather than banning it. Any party using AI in connection with a filing must comply with Rule 11(b) and Rule 26(g), and the court separately prohibits AI-generated legal research that contains false or inaccurate propositions or cites non-existent cases. The rule’s text uses “Artificial Intelligence” without a “generative” qualifier, so it sweeps any AI-assisted work, including non-generative tools. It reaches counsel and unrepresented parties alike.

The prohibition is aimed at the output, not the tool: what is barred is filing AI-generated research that is false, inaccurate, or fabricated, not the act of using AI to prepare a filing. Because the text bars filing false or nonexistent AI-generated citations without prescribing a review or verification step, this entry records the rule as binding without an AI-specific verification facet.

Rule history: this provision was amended

An earlier version of Coleman’s procedures took a markedly stricter line, flatly prohibiting parties from using AI to draft memoranda or to serve as authority supporting motions, with no disclosure-and-verification carveout. That prohibition-style text is the version Ropes & Gray’s tracker captured at 2024-04-03 and the version this entry previously described.

The chambers page now carries the Rule 11(b) / Rule 26(g) formulation quoted below. The page states no adoption date; archived copies show the flat prohibition still posted on June 3, 2026 and the new text in place by late July 2026. Firms working from a copy of this rule saved before mid-2026, or from a secondary summary describing Coleman as a flat-prohibition judge, are working from superseded text.

Ropes & Gray’s tracker lists two 2025 orders in Shanmugavelandy v. AbbVie, before Judge Coleman, as AI-hallucination dispositions. Read that carefully rather than as an enforcement precedent: the plaintiff in that matter was self-represented, the June 2, 2025 order granted summary judgment and terminated the case, the September 10, 2025 order denied a post-judgment motion, and the specific AI findings and any sanction disposition remain unverified against the order text. The tracker holds that matter as an unpublished draft entry for exactly that reason.

Whatever those orders did, they were decided under the prohibition-style version of the chambers rule, so they should be read against the rule text in force on the relevant filing date rather than against the Rule 11 framing that governs now.

Why this matters

The amendment moves Coleman from the strict end of the spectrum toward the mainstream of federal chambers practice. Most chambers rules permit AI use subject to disclosure and verification (the Mehalchick template in M.D. Pa., the Blumenfeld template in C.D. Cal.); several rely on Rule 11 framing without a disclosure mandate (Hodge in E.D. Pa., Robinson in S.D. Cal., Johnston in N.D. Ill.). Coleman’s current rule sits in that second group, alongside Johnston’s, rather than standing alone as an outright ban.

For N.D. Ill. practitioners the practical posture has changed. Firms appearing before Judge Coleman are no longer facing a binary go/no-go on AI tooling; they are facing an accuracy obligation whose failure mode is filing fabricated or inaccurate authority. Documented citation-verification procedures, rather than tool avoidance, are what this rule now tests.

Quotable language

“Any party—counsel and unrepresented parties alike—using Artificial Intelligence (“AI”) in connection with the filing of a pleading, motion, or memorandum of law in this Court must comply with Rule 11(b) and Rule 26(g) of the Federal Rules of Civil Procedure. The court prohibits the use AI-generated legal research containing false or inaccurate propositions or citing non-existent cases.”

Primary source

Judge Coleman’s Case Management Procedures (AI provision), ilnd.uscourts.gov