Artificial intelligence and the legal profession: Practical and ethical considerations in the new technological era

William C. Martucci, who holds an LL.M. from Georgetown University in Washington, D.C., practices nationally in business and employment litigation. He focuses on complex class litigation, including wage and hour litigation, in California, Massachusetts, and New York, as well as business-to-business litigation, executive compensation disputes, business torts, and unfair competition. A nationally regarded trial lawyer, Martucci is often sought out for high-stakes jury work. His jury work has been featured in The National Law Journal.

Burcu Erbaz defends clients in business and employment litigation matters. She graduated first in her class at the University of Missouri-Kansas City School of Law, where she was on the editorial board of the UMKC Law Review. Erbaz also served as a judicial intern for Hon. Stephen R. Bough in the U.S. District Court for the Western District of Missouri, conducting legal research and drafting memoranda.

Minha Jutt represents clients in business litigation matters, including complex commercial disputes, restrictive covenants and trade secrets, and design and construction. Her experience extends to all stages of litigation, ranging from initial client counseling and fact investigation to dispositive motion practice to preparation for appeal. An honors graduate of the University of Kansas School of Law, where she served as the executive comments editor of the Kansas Law Review, Jutt clerked for Hon. Julie A. Robinson, U.S. District Court for the District of Kansas.

The integration of artificial intelligence into the legal profession has rapidly shifted from an experimental novelty to a foundational technological trend, representing in many respects a transformational shift comparable to earlier technological revolutions in the law practice. 

AI systems — especially generative AI tools such as large language models — are being adopted across law firms and courts to support key functions like legal research, drafting, document review, and administrative workflow automation. These technologies can process vast amounts of information and surface relevant statutes, precedents, and legal arguments more quickly than traditional manual methods, ultimately increasing efficiency and enabling legal professionals to prioritize strategic thinking over routine tasks. 

Missouri lawyers now routinely encounter AI in research platforms, document drafting tools, litigation analytics, and client-facing technologies. Used properly, these tools offer meaningful efficiencies, including faster access to information, assistance with initial drafting, and improved access to justice for self-represented litigants. Used improperly, however, generative AI poses serious ethical and professional risks. 

What distinguishes generative AI from earlier tools is not merely speed or convenience, but autonomy. Generative AI systems do not solely retrieve information; they generate content. That distinction is ethically significant. AI-generated outputs are statistical in nature and can produce inaccurate or misleading information, often referred to as “hallucinations.” These hallucinations are not rare anomalies; they are a known limitation of large language models.¹ 

These issues have already drawn judicial ire. As courts across the country have now documented, generative AI can fabricate case law, misstate holdings, and present false information in a manner that appears authoritative.² Several U.S. courts recently sanctioned lawyers for submitting briefs containing fabricated citations derived from unchecked AI output.³ 

Missouri’s ethical legal framework does not prohibit the use of AI. To the contrary, Missouri has long recognized that lawyers must adapt to technological change.⁴ Rule 4-1.1 of the Missouri Rules of Professional Conduct requires competent representation, and Comment [6] expressly includes a duty to stay abreast of “the benefits and risks associated with relevant technology.”⁵ Missouri Informal Opinion 2024-11, issued by the Office of Legal Ethics Counsel, confirms that generative AI is simply the latest technology subject to this longstanding principle.⁶ 

For Missouri lawyers, the ethical issue is not whether AI may be used, but how it may be used consistent with duties of competence, candor, confidentiality, and supervision.⁷ Missouri Informal Opinion 2024-11 properly frames generative AI as analogous to a nonlawyer assistant or external service provider: It’s a tool that may assist the lawyer, but it can never replace the lawyer’s independent judgment or responsibility for accuracy.⁸ 

Using AI in legal practice 

In various practices, perhaps one of the most widely advertised uses of AI is performing traditionally labor-intensive tasks in initial discovery, such as document review and e-discovery. Proponents of AI contend that AI-assisted tools can expedite labor-intensive tasks such as locating key documents, identifying key custodians, and summarizing documents. When performed accurately and thoroughly, these functions can streamline lawyers’ processes and save client expenditures on discovery — ultimately permitting lawyers to expend more time and client funds on more strategic and substantive phases of legal practice. 

Proponents of AI also contend that AI-assisted tools have the potential to streamline legal research by assisting lawyers in quickly locating on-point statutes and case law amidst vast databases. While the accuracy of these tools remains largely untested, they may develop into beneficial aids moving forward. 

AI platforms also have predictive analytic capabilities based on historical data. These capabilities can assess potential case outcomes and settlement ranges, as well as the possibility of the success of a motion based on district and judge-specific data. 

On the transactional side, AI-assisted tools can analyze dense legal terminology to spot inconsistencies, vague or unclear language, and potential risks and unfavorable terms. 

Ultimately, while these developing technologies may streamline and enhance the efficiency of various legal tasks, lawyers remain responsible for ensuring their use of AI complies with their professional obligations. 

Developing guidelines and case law: Missouri ethics and national enforcement trends 

Missouri’s ethics-first approach 

Missouri has addressed the use of AI primarily through ethical interpretation rather than prescriptive court rules. Informal Opinion 2024-11 provides a comprehensive road map for Missouri lawyers considering the use of generative AI.⁹ The opinion emphasizes several core principles:

• Competence (Rule 4-1.1): Lawyers must understand how generative AI works, what it can and cannot reliably do, and the risks associated with its use.¹⁰
• Confidentiality (Rule 4-1.6): Lawyers must evaluate whether AI platforms store, reuse, or expose client data and must make reasonable efforts to prevent unauthorized disclosure.¹¹
• Candor to the tribunal (Rule 4- 3.3): Lawyers may not submit false statements of fact or law, including fabricated citations or mischaracterized authority.¹²
• Supervision (Rules 4-5.1 and 4-5.3): AI output must be supervised just as work performed by nonlawyer assistants, vendors, or junior lawyers is supervised.¹³

The opinion is explicit that lawyers may not rely blindly on AI-generated content. Any work product generated with AI assistance must be independently reviewed and verified by the lawyer before it is used in representation.¹⁴ 

In parallel with these ethical guidelines, Missouri legislators have also proposed statutory guidance over the last couple of years, including in 2025 through HB 1462, the AI Non-Sentience and Responsibility Act.¹⁵ This bill, which did not pass the Missouri Legislature, attempted to codify key principles regarding the legal status and responsibilities associated with AI systems, including:

• Nonsentience: AI systems would have been declared nonsentient and could not hold legal personhood, personal relationships, or corporate roles.
• Human accountability: Owners, developers, and manufacturers would have remained legally responsible for AI outputs, including harm caused through intended or unintended use. Misuse by users would not transfer liability to the AI itself.
• Oversight and safety: Owners would have been required to maintain proper supervision and risk controls over AI systems, with liability for negligence or foreseeable harm. Developers and manufacturers would have to prioritize safety mechanisms and assess risks proactively.
• Piercing corporate veils: The statute made clear that corporate structures could not shield human actors from liability if AI-caused harm resulted from recklessness, negligence, or deceptive practices.

Together, the informal opinion and the proposed legislation reflect a dual approach in Missouri: The ethical guidance ensures lawyers remain competent and responsible in practice, while legislation could establish a clear legal framework assigning human responsibility and preventing AI from being treated as a legal actor. Both underscore that reliance on AI does not absolve lawyers or organizations from accountability. 

Missouri case law: Kruse v. Karlen 

Missouri courts have already enforced these principles. In Kruse v. Karlen, the Missouri Court of Appeals sanctioned a litigant for submitting an appellate brief containing fictitious AI-generated case citations.¹⁶ The court dismissed the appeal and imposed damages, emphasizing that citing nonexistent authority violates fundamental appellate obligations regardless of whether the error arose from AI use.¹⁷ 

Kruse is significant because it demonstrates that Missouri courts do not view AI misuse as a technical lapse, but as a serious breach of professional responsibility.¹⁸ The decision aligns squarely with Informal Opinion 2024-11’s insistence on verification and supervision.¹⁹ 

The federal approach and the push for certification 

Federal courts have likewise responded forcefully to AI-related misconduct. 

Missouri federal courts 

Some judges in the U.S. District Court for the Eastern District of Missouri have individualized standing orders emphasizing the proper use of AI. As an example, Rule 18 of U.S. District Judge Joshua M. Divine’s requirements provides: 

Mandatory Certification Regarding Generative Artificial Intelligence: All attorneys and pro se litigants appearing before the Court must file a certificate on the docket, together with their notice of appearance, attesting either that no portion of any filing will be drafted by generative artificial intelligence (such as ChatGPT, Harvey. AI, or Google Bard) or that any language drafted by generative artificial intelligence will be fully checked for accuracy by a human being, using print reporters or traditional legal databases.²⁰

In addition, the Eastern District of Missouri has issued guidance for self-represented litigants that prohibits filings drafted by any form of AI and emphasizes that litigants — including lawyers — are responsible for the contents of their filings, including portions generated with AI, under Federal Rule of Civil Procedure 11.²¹ 

Kansas federal courts 

Newly promulgated Standing Order 26-01 guides the use of artificial intelligence in preparing court filings in the U.S. District Court for the District of Kansas.²² The order cautions that AI tools may generate false statements of fact or law, including fabricated citations, and reminds litigants that they remain fully responsible for the accuracy of all filings submitted to the court. 

Lawyers and parties must review and verify any AI-assisted content — including legal authority, quotations, factual assertions, and legal analysis — before filing. The order further warns that failure to verify AI-generated material may result in sanctions, including the striking of filings, monetary penalties, referral to disciplinary authorities, disqualification of counsel, filing restrictions, or dismissal of claims. In addition, the court retains discretion to require parties to disclose the use of AI in filings and to certify that any AI-generated material has been personally reviewed and verified for accuracy. 

In a decision on Feb. 2, 2026, District Judge Julie A. Robinson sanctioned multiple lawyers for their improper use of generative AI without independently verifying the legal authority cited in their filings.²³ The court imposed monetary sanctions of $1,000-$5,000 after five lawyers signed briefs containing fabricated and misrepresented case law generated by an AI tool.²⁴ The court emphasized that the use of generative AI in legal practice is not inherently improper; rather, the violation arose from the lawyers’ failure to verify that the cited authorities actually existed and supported the propositions for which they were cited.²⁵ 

As the court explained, the duty to conduct a reasonable inquiry into the law is nondelegable under Federal Rule of Civil Procedure 11, and lawyers who sign filings certify that the legal authorities contained therein have been reviewed and verified through human judgment.²⁶ Robinson further stressed that reasonably competent lawyers should be aware of the well-publicized risks associated with unverified generative AI research, including the potential for hallucinated legal authority, and that reliance on such tools without verification falls short of the objective standard required by Rule 11.²⁷ 

In Moore v. City of Del City,²⁸ the 10th U.S. Circuit Court of Appeals dismissed an appeal as a sanction after a pro se litigant cited numerous fabricated cases that the court determined were generated by AI.²⁹ The court emphasized that failure to verify citations wastes judicial resources and undermines confidence in the legal system.³⁰ Importantly, the 10th Circuit imposed a prospective certification requirement, ordering the litigant to disclose future AI use and verify citation accuracy under penalty of perjury.³¹

While Missouri has not adopted a formal certification rule, a functional equivalent already exists. Every Missouri lawyer who signs a pleading or brief certifies — by signature — that the authority cited exists and supports the propositions asserted.³² The use of AI does not alter that obligation. 

National cases 

In Mata v. Avianca, Inc., lawyers relied on ChatGPT to draft a motion that cited six nonexistent judicial decisions.³³ The U.S. District Court for the Southern District of New York found that counsel failed to verify the AI-generated authorities before filing and imposed $5,000 in sanctions, while also ordering the lawyers to notify the judges whose names had been falsely attributed to fabricated opinions.³⁴ Widely regarded as the first major sanctions decision addressing generative AI in federal litigation, Mata established the now-frequently cited principle that the use of AI does not excuse violations of Rule 11.³⁵ 

More recently, in February 2026, the 5th U.S. Circuit Court of Appeals reaffirmed that traditional sanction rules adequately govern AI-related misconduct. In Fletcher v. Experian Information Solutions, Inc., the court observed that although generative AI represents a new technology, existing procedural rules already provide sufficient mechanisms to address inaccurate or fabricated legal citations.³⁶ The court emphasized that the same professional and procedural obligations apply regardless of whether legal research is performed by traditional means or assisted by AI.³⁷ 

Similarly, in Lifetime Well LLC v. Ibspot.com, the U.S. District Court for the Eastern District of Pennsylvania highlighted the supervisory responsibilities of lawyers when AI tools are used in preparing filings.³⁸ The court warned that a lack of diligence and supervision — particularly where less experienced lawyers rely on generative AI without adequate review — may lead to both monetary and nonmonetary sanctions when filings contain false or fabricated legal authority.³⁹ 

Together, these cases reinforce a consistent judicial message: While AI may assist with legal research and drafting, the responsibility for verifying legal authority and ensuring the accuracy of court filings remains squarely with the lawyer. 

The American Bar Association 

The American Bar Association issued formal ethical guidance on lawyers’ use of AI in ABA Formal Opinion 512 and through interpretations of the ABA Model Rules of Professional Conduct. These authorities do not prohibit the use of AI in legal practice but instead require that lawyers employ such tools in a manner consistent with their professional obligations. 

Formal Opinion 512 explains that existing ethical rules fully apply when lawyers use AI-assisted tools.⁴⁰ The duties of competence, confidentiality, supervision, candor to the tribunal, and reasonable fee practices remain the responsibility of the lawyer.⁴¹ Lawyers must independently review and verify AI-generated outputs before relying on them in legal work and must take care to avoid entering confidential client information into publicly accessible AI systems.⁴² 

The opinion further encourages law firms to implement internal policies and supervisory measures governing the use of AI.⁴³ In addition, lawyers must ensure that court filings contain verified legal authority and may need to disclose the use of AI to clients when it materially affects legal strategy, confidentiality, or costs.⁴⁴ 

In essence, once again, the responsibility for the final legal work product rests with the lawyer — not with the AI system — and lawyers remain accountable for verifying the accuracy and integrity of any AI-assisted work.

Judicial AI misuse 

A recent Bloomberg Law article reported that two federal judges issued erroneous rulings after chambers staff used generative AI tools without adequate review.⁴⁵ While these incidents sparked concern about judicial oversight, their relevance to Missouri lawyers lies elsewhere: They reinforce that accountability follows authority.⁴⁶ Just as judges remain responsible for opinions issued under their names, Missouri lawyers remain responsible for filings submitted under theirs. 

Considering AI in practice 

Proper training 

While AI-assisted tools continue to be developed, they are not infallible. Law firms should provide training to their lawyer and nonlawyer employees about the ethical and practical considerations to be made regarding the use of AI. Law firms should also establish clear policies regarding permissible uses of AI. 

Lawyers must familiarize themselves with the extent of the capabilities and limits of the AI-assisted tools. Just as law students and lawyers traditionally underwent training for Westlaw and LexisNexis, they should similarly undergo training for each AI-assisted tool they intend to deploy. A lawyer should be able to understand and articulate the mechanism an AI tool uses in its analysis, especially when that analysis influences an important strategic decision. 

Beyond attending vendor-led trainings, lawyers should independently explore and test AI-assisted tools to ensure they can fulfill their intended purposes. For example, a lawyer may ask an AI-assisted tool to review and summarize an email or contract the lawyer has already reviewed before deploying the tool more widely. 

Confidentiality

Lawyers must ensure that AI-assisted tools preserve the confidentiality of the information — both client confidentiality and any information marked as confidential through a protective order or similar mechanism. Under Missouri law, lawyers must make reasonable efforts to protect client information. In the context of AI, law firms and lawyers should ensure all AI systems used are “closed,” meaning the systems store information in protected databases. Law firms should consult with IT or cybersecurity professionals to assess risks associated with security, data privacy, and unauthorized data retention. Lawyers should review applicable terms of use for each platform to assess the risk of unauthorized disclosure. 

Lawyers should further advise their clients regarding the use, risks, and benefits of AI before obtaining informed consent from the client if necessary. 

Supervision 

All work products generated with AI assistance should be independently reviewed and verified by a lawyer before the work product is used in client matters. As numerous case law and anecdotal examples demonstrate, AI is not infallible. Lawyers should scrutinize all AI-generated work product for accuracy, bias, and thoroughness and correct any deficiencies. 

Over-reliance 

Lawyers should not over-rely on AI-assisted tools for tasks such as legal research, contract analysis, or predictive capabilities. Instead, lawyers should assess case needs, client goals, varying strategies, possible outcomes, and risks to make an informed decision regarding the best use of AI to supplement the lawyer’s own analysis rather than replace it. 

Looking ahead: Practical and ethical implications for Missouri lawyers 

Missouri is entering a phase of normalization of AI usage coupled with heightened accountability. AI will remain embedded in legal practice, but tolerance for careless or unsupervised use is rapidly disappearing. For Missouri practitioners, the road map is clear. 

First, courts are increasingly unwilling to accept ignorance, delegation, or automation as excuses for inaccurate filings. Therefore, verification must become routine. Missouri lawyers should assume courts will expect every citation, quotation, and legal proposition — whether AI-assisted or not — to be independently confirmed. This expectation is not merely aspirational, it is increasingly being enforced, and the risk of sanctions is real. As Kruse and Moore demonstrate, courts are expanding sanctions beyond monetary penalties and are willing to dismiss cases, strike filings, and impose prospective restrictions when AI misuse results in false authority or misrepresentations of the law. 

Second, law firms must address supervision explicitly. Verification obligations apply regardless of who uses the AI — lawyer, intern, nonlawyer assistant, or law clerk. Informal Opinion 2024-11 makes clear that partners and managers bear responsibility for ensuring that lawyers and nonlawyers understand appropriate AI use. 

Third, confidentiality analysis is essential. Missouri lawyers must scrutinize AI platform terms of service and data-handling practices, particularly where client information is involved. 

Finally, Missouri lawyers should resist the false choice between innovation and ethics. Banning AI outright is neither realistic nor desirable. The ethical path forward is informed, careful, and accountable use. 

Looking forward, Missouri courts may consider whether to formalize AI practices through standing orders, appellate rules, or administrative guidance, particularly as federal courts experiment with disclosure and certification requirements. Even absent such rules, Missouri’s existing Rules of Professional Conduct and Missouri Informal Opinion 2024-11 already supply a workable ethical framework: Educate yourself on the technology, safeguard confidentiality, supervise its use, verify its output, and never abdicate professional judgment. 

Generative AI does not change the lawyer’s role as gatekeeper, advocate, and officer of the court. It amplifies that role. It is a powerful tool that magnifies both competence and carelessness. For Missouri lawyers, the ethical question is not whether AI will be used, but whether it will be used in a way that preserves the competence, candor, and integrity of the profession.⁴⁷

Endnotes 
1 What are AI Hallucinations?, Google Cloud, https://cloud.google.com/discover/what-are-ai-hallucinations (last visited Dec. 30, 2025). 
2 Moore v. City of Del City, No. 25-6002, 2025 WL 3471341, at *2 (10th Cir. Dec. 3, 2025); see also Wadsworth v. Walmart Inc., 348 F.R.D. 489, 497 (D. Wyo. 2025) (“It is ... well-known in the legal community that AI resources generate fake cases.”). 
3 Daniel Wu, Lawyers Using AI Keep Citing Fake Cases in Court. Judges Aren’t Happy, WASH. POST (June 3, 2025), https://www.washingtonpost.com/nation/2025/06/03/attorneys-court-ai-hallucinations-judges/. 
4 MO. RULES OF PRO. CONDUCT R. 4-1.1 cmt. [6] (2025). 
5 Id.
6 MO. INFORMAL OP. 2024-11. 
7 MO. INFORMAL OP. 2024-11, supra note 4. 
8 Id. 
9 Id. 
10 Id.; MO. RULES OF PRO. CONDUCT R. 4-1.1. 
11 Id.; MO. RULES OF PRO. CONDUCT R. 4-1.6. 
12 Id.; MO. RULES OF PRO. CONDUCT R. 4-3.3. 
13 Id.; MO. RULES OF PRO. CONDUCT R. 4-5.1 and 4-5.3. 
14 MO. INFORMAL OP. 2024-11, supra note 4. 
15 AI Non-Sentience and Responsibility Act, H.B. 1462, 101st Gen. Assemb., Mo. (2025), available at https://legiscan.com/MO/text/HB1462/id/3141592 (last visited Mar. 16, 2026). 
16 Kruse v. Karlen, 692 S.W.3d 43, 49 (Mo. Ct. App. 2024), reh’g and/or transfer denied (Apr. 9, 2024). 
17 Id. at 53–54. 
18 Id.
19 MO. INFORMAL OP. 2024-11, supra note 4.
20 Requirements, Hon. Joshua M. Divine, U.S. Dist. Ct. for the E.D. Mo. (revised August 2025), available at https://www.moep.uscourts.gov/judge/joshua-m-divine. 
21 See Guidance for Self-Represented Litigants on the Use of Generative Artificial Intelligence (United States District Court for the Eastern District of Missouri), available at https://www. moed.uscourts.gov/self-represented-litigants-srl. 
22 U.S. Dist. Ct. for the Dist. of Kan., Standing Order No. 26-01 (2026), available at https://www.ksd.uscourts.gov/news/notice-proposed-amendment-local-rules-issuance-so-26-01-and-changes-form-protective-order (last visited Mar. 16, 2026). 
23 Doc. 218, Memorandum and Order, at 1, Overstock.com, Inc. v. Lexos Media IP, LLC, No. 2:22-cv-02324-JAR-ADM (D. Kan. Feb. 2, 2026).
24 Id. at 35–36. 
25 Id. at 20–21. 
26 Id. at 19–21. 
27 Id. at 26–28.
28 25-6002 (10th Cir. Dec. 03, 2025).
29 Moore, 2025 WL 3471341, at *3, supra note 5. 
30 Id. at *2.
31 Id. at *3.
32 MO. RULES OF PRO. CONDUCT R. 4-3.1.
33 Mata v. Avianca, Inc., 678 Fed. Supp. 3d 443, 451–52, 465–66 (S.D.N.Y. 2023). 
34 Id. 
35 Id. at 461–62 (explaining that fabricated authorities generated by AI cannot constitute "existing law” for purposes of Rule 11); see also Wadsworth v. Walmart Inc., 348 F.R.D. 489, 493 (D. Wyo. 2025) (sanctioning attorneys under Rule 11 for filing a motion that cited multiple nonexistent cases generated by AI). 
36 Fletcher v. Experian Info. Sols., Inc., No. 25-20086, 2026 WL 456842, at *3 (5th Cir. Feb. 18, 2026). 
37 Id. at *5–6. 
38 Lifetime Well, LLC v. Ibspot.com, Inc., No. 23-cv-02203, 2026 WL 195644, at *1 (E.D. Pa. Jan. 26, 2026). 
39 Id. at *4–5. 
40 See ABA Formal Op. 512 (2024). 
41 Id. at 2–4, 7–10 (2024). 
42 Id. at 5–6. 
43 Id. at 10. 
44 Id. at 7–9. 
45 Justin Henry, Judges Admit to Using AI After MadeUp Rulings Called Out, BLOOMBERG LAW (Oct. 23, 2025), https://news.bloomberglaw. com/business-and-practice/judges-called-out-for-nonfactual-rulings-admit-to-use-of-ai.
46 Stephen Gillers (Prof., NYU Law), quoted in Judges Admit to Using AI After MadeUp Rulings Called Out, BLOOMBERG LAW, supra note 47 (discussing the need for judges and lawyers to verify AI-generated citations); For additional insights regarding AI in this context of judicial engagement, see generally Anika Jaitley, Daniel W. Linna Jr., Hon. Xavier Rodriguez, V.S. Subrahmanian & Siyu Tao, Artificial Intelligence in Federal Courts: A Random-Sample Survey of Judges, 27 SEDONA CONF. J. __ (forthcoming 2026). 
47 See generally New York State Bar Association, Report and Recommendations of the New York State Bar Association Task Force on Artificial Intelligence (April 2024). “Among the many comments and insights offered in connection with the New York State Bar Association Task Force on Artificial Intelligence, the conclusion is instructive in the insights provided. This report offers no “conclusions.” As AI continues to evolve, so will the work of NYSBA and the groups tasked with ongoing monitoring. As a profession, lawyers and judges must continue to refine the initial guidelines suggested in this report and audit the efficacy of proposed rules and regulations. Legal professionals are encouraged to liken this journey to the mindset of ancient explorers: Be cautious, be curious, be vigilant, and be brave.”