June 1, 2026 (in 3 days): New York: 22 NYCRR Part 161 takes effect, system-wide AI policy for all UCS courts

Johnson v. Dunn

U.S. District Court, Northern District of Alabama · N.D. Ala. · Alabama bar guidance

Court sanction

Verified May 5, 2026

Citation
Johnson v. Dunn, 792 F.Supp.3d 1241 (N.D. Ala. 2025)
Decided
July 23, 2025

Summary

Three Butler Snow LLP attorneys -- partners Matthew B. Reeves, William J. Cranford, and William R. Lunsford -- filed two discovery motions on behalf of former Alabama Department of Corrections Commissioner Jefferson Dunn in a prisoner civil rights case. The motions contained five fabricated case citations. Partner Reeves had used ChatGPT to generate legal research and did not verify the citations before filing. Butler Snow had adopted a written firm AI use policy in June 2023 that required written approval and independent verification; the policy was not followed.

AI tool:
ChatGPT
This case summary is informational only. Verify the underlying opinion or order against the primary source before relying on it in any filing or client matter.

What sanction did the court impose?

U.S. District Judge Anna Manasco disqualified all three attorneys from representing the defendant for the remainder of the case, directed that the opinion be published in the Federal Supplement, and referred the attorneys to bar regulators in every state where they hold licenses. Sanctioned attorneys were required to disclose the sanctions order to all current clients, opposing counsel, and presiding judges in all pending matters. The court expressly rejected monetary sanctions as insufficiently deterrent: 'If fines and public embarrassment were effective deterrents, there would not be so many cases to cite.' Butler Snow itself was not sanctioned, in recognition of its pre-existing written AI use policy and cooperative response.

Why does Johnson v. Dunn matter for law firms using AI?

Johnson v. Dunn is the case where monetary sanctions were rejected as too lenient. Judge Manasco disqualified three partners from a national firm and referred them to every state bar where they are licensed, reasoning that fines have not deterred the pattern. Two other points make the case distinctive: Butler Snow itself escaped sanction because it had a written AI policy (the partner just had not followed it), and the Federal Supplement publication was intentional — the court wanted the order to carry precedential weight for the next hallucination case that reaches a federal judge’s desk.

Sources

Primary sources

Further reading