United States Bankruptcy Court for the Northern District of Illinois, July 18, 2025
On July 18, 2025, the United States Bankruptcy Court for the Northern District of Illinois (Eastern Division), presided over by the Honorable Michael B. Slade, issued a significant memorandum opinion in the case In re Marla C. Martin, Case No. 24 B 13368.
In this decision, the court found a violation of Federal Rule of Bankruptcy Procedure 9011 by debtor’s counsel for submitting a legal brief containing fabricated case citations generated by artificial intelligence, without verification.
The ruling reflects a growing judicial consensus across the United States on the ethical and professional risks associated with the uncritical use of generative AI tools in legal practice.
The case concerned a Chapter 13 bankruptcy proceeding initiated by Marla C. Martin, who was represented by The Semrad Law Firm, LLC. It was the debtor’s eighth bankruptcy case, with each of her previous filings dismissed before completion. The current proceeding involved substantial arrears in real estate taxes on her Chicago residence, giving rise to a tax lien held by Corona Investments, LLC, which objected to the debtor’s proposed Chapter 13 plan.
In response to Corona’s objection, debtor’s counsel filed a brief (Dkt. No. 51) challenging the creditor’s standing to raise certain issues, notably the feasibility of the plan under 11 U.S.C. § 1325(a)(6). The brief cited four purported decisions, In re Montoya, In re Jager, In re Coleman, and In re Russell, which were either misrepresented or entirely fictitious. The quotations attributed to these decisions did not appear in the actual opinions, and two of the cases either did not exist or were misattributed in terms of jurisdiction and citation.
When questioned at the June 10, 2025 hearing, attorney Thomas E. Nield admitted that he had used artificial intelligence, specifically ChatGPT, to generate portions of the legal argument and citations, without verifying the existence or substance of the cases referenced.
The court held that the conduct of Mr. Nield and The Semrad Law Firm, LLC constituted a clear violation of Rule 9011(b)(2) of the Federal Rules of Bankruptcy Procedure, which requires attorneys to ensure, after reasonable inquiry, that legal contentions presented to the court are warranted by existing law or by a non-frivolous argument for the extension or modification of the law.
Citing a growing body of recent federal decisions, including Mata v. Avianca, Mid Central Operating Engineers Health & Welfare Fund v. HoosierVac LLC, and Benjamin v. Costco Wholesale Corp., the court emphasized that attorneys must verify the accuracy and validity of any legal authorities cited in their filings. Generative AI tools, the court noted, are well known to “hallucinate” citations and fabricate nonexistent precedent.
Judge Slade rejected the assertion that ignorance of the limitations of AI could excuse such conduct. He wrote: “At this point, to be blunt, any lawyer unaware that using generative AI platforms to do legal research is playing with fire is living in a cloud.” The court further recalled that the Illinois Supreme Court adopted, effective January 1, 2025, a formal policy on artificial intelligence requiring practitioners to “thoroughly review AI-generated content before submitting it in any court proceeding.”
Taking into account counsel’s candor, the withdrawal of the offending brief, the firm’s withdrawal of its fee application under 11 U.S.C. § 329, and the internal steps taken by the law firm after the incident (including the adoption of a formal AI policy and CLE training), the court imposed what it described as the “least harsh sanction” necessary to deter similar conduct.
Specifically, the court ordered:
The court expressly declined to impose further sanctions such as an ARDC referral or a Section 329 forfeiture, noting that Mr. Nield had already self-reported his conduct to disciplinary authorities and that the firm had withdrawn its fee request.
This decision contributes to a growing jurisprudential framework addressing the misuse of generative AI in legal advocacy. It reaffirms that AI-generated text, no matter how convincingly phrased, must never substitute for a lawyer’s independent judgment, due diligence, and verification. The court emphasized that “no lawyer should be using ChatGPT or any other generative AI product to perform research without verifying the results. Period.”
As the legal profession confronts the accelerating integration of AI technologies into practice, this ruling stands as a warning: failing to supervise the outputs of such tools may result not only in professional embarrassment but in formal sanctions, monetary penalties, and reputational harm.
This article was prepared by a lawyer admitted to practice in France. Any interpretation or legal opinion concerning U.S. law should be conducted in consultation with an attorney duly licensed in the United States.
