Cautions for Using Generative AI in the Legal Profession
Cautions for Using Generative AI in the Legal Profession
Like other tools, artificial intelligence can enhance a lawyer’s practice by simplifying research and drafting tasks. But at the same time, any lawyer that cedes his or her legal work entirely to any AI tool violates not only the rules of professional conduct but also risks permanent reputational damage. Over the last few years, there have been numerous widely discussed instances where attorneys filed legal briefs—written in whole or in part by AI tools—that were later found to be riddled with errors, including misrepresentations of facts and citations to non-existent authorities fabricated by the AI tool. And even if not exposed, a lawyer’s use of AI without sufficient verification and substantiation may jeopardize the accuracy of legal advice the lawyer provides to the client.
A Lawyer’s Ethical Obligations in Using AI
Regardless of what tools a lawyer chooses to employ in representing his or her client, it is the lawyer who is licensed and has undertaken the solemn obligation to represent clients ethically in accordance with the applicable rules of professional conduct. When dealing with new and emerging technologies, the Professional Ethics Committee of the Texas State Bar has consistently held that a lawyer who employs a technological tool must “have a reasonable and current understanding of the technology.”[2]
Specifically with respect to generative AI tools, the Texas Bar has stated that only by having such an understanding “can the lawyer evaluate the associated risks of hallucinations or inaccurate answers, the limitations that may be imposed by the [AI tool]’s use of incomplete or inaccurate data, and the potential for exposing client confidential information.”[3] The obligation of competence therefore impacts other ethical rules, including Rule 1.05, which addresses protecting confidential client information.[4] The rules directed to a lawyer’s responsibility when representing a client as an advocate, such as Rules 3.01 and 3.03, which speak to a lawyer’s obligation to take only meritorious positions in a proceeding and to avoid presenting false statements of material fact or law to a tribunal, may also be implicated.[5]
Accuracy of AI Generated Materials
Numerous cases publicized in the legal and general media have highlighted one of the most significant risks of using generative AI tools for drafting court documents—the (very real) possibility that the AI tool will hallucinate and include in the generated document citations to authorities that do not exist or that do not stand for the propositions cited. Just recently, a (now former) attorney for a man sued by renowned Dallas pastor T.D. Jakes was sanctioned by a Pennsylvania federal judge for submitting court filings containing “misrepresentations and fabrications of relevant case law.”[6] According to the judge, “[a]t this point, any attorney—but especially attorneys like [the sanctioned lawyer] with a cosmopolitan, multistate, practice—cannot plead ignorance to justify their use or misuse of AI in drafting legal documents.”[7]
In another high-profile example from 2023, President Trump’s former attorney, Michael Cohen, landed his lawyer in hot water by reviewing the lawyer’s draft brief of a motion for early release from prison and then sending the lawyer proposed edits with a few case citations Cohen had found using Google Bard, a generative AI tool. Unbeknownst to Cohen—and to the detriment of his lawyer who failed to check the citations before including them in Mr. Cohen’s motion for early release—the cited caselaw had been fabricated by the AI tool.[8] Ultimately, the court determined that sanctions were not warranted under the applicable standards because the false citations were not made in “bad faith,” but the court nevertheless chided the attorney, stating that the “citation to non-existent cases is embarrassing and certainly negligent, perhaps even grossly negligent.”[9]
These cases, however, are just the tip of the iceberg, as dozens of lawyers across the country have found themselves on the wrong side of an angry judge after submitting inaccurate briefs generated by AI with little or no verification of the briefs’ accuracy. In Texas alone, there are at least fifteen cases where a lawyer has faced sanctions for submitting inaccurate briefing to federal or state courts where the source of the inaccuracies was a generative AI tool.
How States and Courts Have Addressed the Problem
Bar associations and courts around the country recognize that the improper use of generative AI tools is a serious problem that has the potential to undermine faith in the judicial system. There is, however, no consistency in how best to address the problem. Thus, what is expected and required of lawyers is found in a patchwork of rules and ethics opinions from state bar associations, state- or district-wide court rules, or even standing orders from particular judges. And these various requirements may impose differing restrictions, making it even more imperative that the lawyer be aware of how and to what extent he or she may use AI tools in the various jurisdictions in which they practice.
As discussed, the Texas Bar Ethics Committee has issued an opinion regarding the use of generative AI tools. That opinion warns of the risks of using AI and provides practical—albeit obvious—guidance that a lawyer “should always verify the accuracy of any responses received from a generative AI tool,” lest they “risk many of the same perils as those who rely on inexperienced or overconfident nonlawyer assistants.”[10] The Texas opinion stops short, however, of requiring the lawyer to obtain client consent for the use of AI tools. Instead, it merely states that “[i]f the lawyer intends to use confidential information in conjunction with generative AI tools, the lawyer should consider informing clients about the associated risks and may need to secure client consent.”[11] The opinion provides no suggestion about obtaining client consent, or even disclosing an intent to use generative AI, in situations that do not involve the use of confidential information.
Beyond Texas, a handful of states have adopted guidelines indicating the need for a lawyer to obtain the client’s consent for using generative AI tools in certain contexts.[12] Other states have yet to issue guidance or are silent as to whether client consent is required. On a more granular level, courts and specific judges are imposing requirements for lawyers and pro se litigants to disclose the use of AI tools to the court. In Texas, for example, federal and state courts have adopted rules concerning AI, and some judges have issued standing orders for cases in their courts. These range from simple reinforcements that a lawyer’s obligations are still present even when using AI tools[13] to requirements that lawyers certify whether any generative AI tools were used in the preparation of materials submitted to the court.[14]
Practical Implications
Lawyers choosing to employ AI tools should do so cautiously and ensure that they continue to abide by all ethical and court rules and requirements. If using an open-type AI tool such as ChatGPT or Google Gemini, the lawyer should refrain from inputting any confidential information into the tool. But if the lawyer believes that doing so is important to the representation, he or she should first obtain the client’s informed, written consent to minimize any misunderstandings.
Although there has been considerable academic research into reducing hallucinations and the generation of factually incorrect information in generative AI tools,[15] they cannot be eliminated, especially where open-type tools are used for legal work. Thus, a lawyer using generative AI for legal drafting must be extra vigilant about verifying the accuracy of all factual details and caselaw in AI-generated documents. And if practicing in a court where the rules or standing orders require the lawyer to certify whether AI was used to draft the submission in whole or in part, a lawyer may want to consider the optics of using an AI tool at all.
Similar concerns are also present when using generative AI to create other types of legal documents, including contracts. Generative AI tools may import into a draft contract provisions that may be irrelevant or improper for the particular contract the lawyer is drafting. Thus, the lawyer must exercise care to ensure that any extraneous or improper provisions are corrected before the contract is finalized.
The bottom line is that a lawyer should understand enough about AI and how it works to make the decision whether to use it. And if the lawyer does so, he or she needs to ensure compliance with all applicable ethics requirements and any court rules or standing orders. Most importantly, the lawyer should presume that any documents generated by an AI tool will contain errors, and thus, should check and verify everything in the document to confirm its accuracy.
[1] No portion of this article was generated with artificial intelligence.
[2] Tex. Bar Prof. Ethics Comm. Op. No. 705, at 2 (Feb. 2025); see also Tex. Bar Prof. Ethics Comm. Op. No. 680, at 3 (Sept. 2018) (reiterating that Rule 1.01(a) of the Texas Disciplinary Rules of Professional Conduct impose a “competency” requirement for attorneys that is applicable to the lawyer’s technological competence in using technological tools in conjunction with the practice of law).
[3] Tex. Bar Prof. Ethics Comm. Op. No. 705, at 2.
[4] See Tex. Disciplinary R. Prof. Conduct 1.05.
[5] See Tex. Disc. R. Prof. Conduct 3.01, 3.03.
[6] T.D. Jakes Case, Lawyer fined over AI, Dallas Morning News, at B–1 (Oct. 11, 2025).
[7] Jakes v. Youngblood, No. 2:24cv1608, slip op. at 2 (W.D. Pa. Oct. 6, 2025).
[8] U.S. v. Cohen, No. 18 Cr. 602, Ltr. from E.D. Perry to Court [ECF 104], at 1 (S.D.N.Y. Dec. 29, 2023).
[9] U.S. v. Cohen, No. 18 Cr. 602, Opinion and Order of Mar. 20, 2024, slip op. at 11 (S.D.N.Y. Mar. 20, 2024).
[10] Tex. Bar Prof. Ethics Comm. Op. No. 705, at 5.
[11] Id. at 4 (emphasis added).
[12] See Fla. Bar Ethics Op. 24–1 (Jan. 19, 2024) (recommending obtaining client consent before using AI tools that utilize client confidential information); Ill. State Bar. Ass’n Report § IV.B (Sept. 27, 2023) (recommending rule to require client consent before using generative AI tools); Mich. Bar Ass’n Frequently Asked Questions (Nov. 18, 2024) (client consent is required only where the use of an AI “program may expose confidential information”); N.C. State Bar, 2024 Formal Ethics Op. 1 (Nov. 1, 2024) (client consent is required for a lawyer to delegate “substantive tasks in furtherance of the representation to an AI tool”); W. Va. Legal Ethics Op. 24–01 (June 14, 2024) (informed, written consent is required before a lawyer may use generative AI).
[13] See, e.g., E.D. Tex. Local Rule AT-3(m); S.D. Tex. Gen. Order 2025-04 (May 7, 2025); Tex. Bus. Ct. Local Rule 10(c); .
[14] See, e.g., N.D. Tex. Local Rule 7.2(f); N.D. Tex. Bankr. Gen. Order 2023–03 (June 21, 2023); Bexar County Civ. Dist. Cts. Local Rule 3.H; Standing Order Regarding Use of Artificial Intelligence (Tex. 109th Dist. Ct. Dec. 6, 2024); Standing Order Regarding Use of Artificial Intelligence (Tex. 30th Dist. Ct. Mar. 26, 2024).
[15] See, e.g., Orion Weller et al., “According to . . .”: Prompting language Models Improves Quoting from Pre-Training Data, Johns Hopkins Univ. (Feb. 26, 2024); Long Ouyang et al., Training Language Models to Follow Instructions with Human Feedback, OpenAI (Mar. 4, 2022); Ziwei Ji et al., Survey of Hallucination in Natural Language Generation, ACM Comput. Survey, Vol. 1, No. 1 (Feb. 2022).
Klemchuk PLLC is a leading IP law firm based in Dallas, Texas, focusing on litigation, anti-counterfeiting, trademarks, patents, and business law. Our experienced attorneys assist clients in safeguarding innovation and expanding market share through strategic investments in intellectual property.
This article is provided for informational purposes only and does not constitute legal advice. For guidance on specific legal matters under federal, state, or local laws, please consult with our IP Lawyers.
© 2025 Klemchuk PLLC | Explore our services