Is an empty head, but pure heart a legal defense to using generative artificial intelligence to do your work? For attorneys, the answer has been and continues to be, no. Two recent cases illustrate why.

In February 2025, U.S. Magistrate Judge Mark Dinsmore for the U.S District Court for the Southern District of Indiana in Mid Central Operating Engineers Health and Welfare Fund v. HoosierVac, LLC, recommended that Rafael Ramirez, a Texas attorney, be sanctioned $15,000 for relying entirely on generative AI to perform legal research. After the Magistrate Judge was unable to find a case that Ramirez had cited in a legal brief filed with the Court, Ramirez withdrew the citation to the case and apologized to the Court and opposing counsel. The Magistrate Judge then undertook a non-exhaustive review of two prior legal briefs that Ramirez had filed in the same case and found additional citations to cases that simply did not exist.

At a ‘show cause’ hearing that the Magistrate Court scheduled, Ramirez explained that “he had used AI before to assist with legal matters, such as drafting agreements, and did not know that AI was capable of generating fictitious cases and citations.” Since Ramirez deemed that the ‘hallucination cites’ appeared credible, he did not conduct any further research or make any attempt to verify the existence of the generated citations.

Although Ramirez pleaded with the Court that he did not act in bad faith or proceed with malice, the Court found that immaterial to Ramirez’s obligations as an attorney. Relying on Rule 11 of the Federal Rule of Civil Procedure, the Court examined whether Ramirez’s use of generative AI to frame his legal arguments was objectively reasonable under the circumstances. Quoting from a to a 1986 Seventh Circuit Court of Appeals case, a case decided before the widespread use of AI,  the Magistrate Judge noted:

Rule 11 requires counsel to study the law before representing its contents to a federal court. An empty head but a pure heart is no defense.

The Magistrate Judge found that Ramirez’s “failing to check the treatment and soundness—let alone the existence—of a case warrants sanctions” and recommended sanctions of $5,000 per brief that contained ‘hallucinogenic cites’, of $15,000 total. However, the Judge did not stop  there. Finding that Ramirez violated three separate Indiana Rules of Professional Conduct, the Judge recommended that Ramirez’s transgressions be referred to the appropriate authorities for disciplinary proceedings. If those authorities find Ramirez acted unprofessionally, it may have ramifications in other states in which Ramirez is licensed to practice, including Texas.

Across the country, but also in February 2025, the U.S. District Court Judge for Wyoming in Wadsworth v. Walmart, Inc. et al., issued an order for monetary sanctions against Morgan & Morgan for including hallucinogenic cites in motions in limine. As the Judge explained, a “hallucination occurs when an AI database generates fake sources of information.” Relying on a secondary source, the Court explained how hallucinogenic citations occur:

AI models are trained on data, and they learn to make predictions by finding patterns in the data. However, the accuracy of these predictions often depends on the quality and completeness of the training data. If the training data is incomplete, biased, or otherwise flawed, the AI model may learn incorrect patterns, leading to inaccurate predictions or hallucinations.

Unlike many of the prior sanctions matters, including the one against Ramirez, addressed above, Morgan & Morgan used a proprietary, in-house AI program rather than a public one, like ChatGPT, to perform legal research. In particular, a Morgan & Morgan attorney uploaded the brief for the motions in limine into “MX2.law” to add case law, making the following requests within the program:

1.   “Add to this Motion in Limine Federal Case law from Wyoming setting forth requirements for motions in limine”
2.   “Add more case law regarding motions in limine”
3.   “Add a paragraph to this motion in limine that evidence or commentary regarding an improperly discarded cigarette starting the fire must be precluded because there is no actual evidence of this, and that amounts to an impermissible stacking of inferences and pure speculation.”

The search inquiries generated citations to fake cases, and without verifying their accuracy, Morgan & Morgan included the fake cases in the motions in limine. The two attorneys who signed the motions in limine argued they relied on a third attorney to perform the research and trusted that the research was accurate. Morgan & Morgan claimed to first learn that the cases were questionable when the Court entered the order to show cause.

The Court found that Morgan & Morgan’s use of fake cases violated Rule 11, and provided the following guidance to any attorney signing any pleading.

The key takeaway for attorneys is simple: make a reasonable inquiry into the law before signing (or giving another permission to sign) a document, as required by Rule 11. If an attorney does not do so, then they should not sign the document. However, if the attorney decides to risk not making reasonable inquiry into the existing law and signs, then they may be subject to sanctions.

Unlike the Ramirez case, the Court ordered lesser sanctions against Morgan & Morgan, in part because at the time of the show cause hearing, Morgan & Morgan had already withdrawn its motion in limine, fessed up to using AI, paid opposing counsel’s attorney’s fees for defending against the motion, and implemented policies to present another future occurrence. The Court ordered that the attorney who relied on the AI program to perform the research be sanctioned $3,000 and the two attorneys who signed the pleading be sanctioned $1,000 each.

The attorneys in our Austin and Dallas office are frequently asked questions by clients and opposing counsel about the use of AI to prepare documents, perform support legal services, and assist in legal matters. If you should have any questions, please contact us at info@gstexlaw.com.

Legal Disclaimers

This blog is made available by Gerstle Snelson, LLP for educational purposes and to provide general information about the law, only.  Neither this document nor the information contained in it is intended to constitute legal advice on any specific matter or of a general nature.  Use of the blog does not create an attorney-client relationship with Gerstle Snelson, LLP where one does not already exist with the firm.  This blog should not be used as a substitute for competent legal advice from a licensed attorney.

©Gerstle Snelson, LLP 2025.  All rights reserved.  Any unauthorized reprint or use of this material is prohibited.  No part of this blog may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or by any information storage or retrieval system without the express written permission of Gerstle Snelson, LLP.