I. INTRODUCTION

Several months ago, a friend of mine introduced a speaker at a meeting of trial lawyers. He told the audience that he had used artificial intelligence (“AI”) to prepare the introduction, and he then read it to us verbatim. It was a beautiful introduction. There was, however, one serious problem. Certain quotes included were completely fabricated, as the introducer frankly acknowledged. The lesson we learned that day was, if you use AI in your law practice, it is critical you be aware of both the risks inherent in its use, as well as its effect on your ethical obligations.

Plainly, AI is not ready to replace the lawyer’s professional judgment, as a recent survey of Texas lawyers confirms. A 2024 survey by the State Bar of Texas Law Practice Management Program indicated that 67% of responders do not use AI at all in their law practice, and 49% do not use AI because of ethical concerns.

II. AI AND GENERATIVE AI

The National Artificial Intelligence Initiative Act of 2020 defines “artificial intelligence” as “a machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations or decisions influencing real or virtual environments; abstract such perceptions into models through analysis in an automated manner; and use model inference to formulate options for information or action.” To be sure, AI technology continues to advance, and AI is now capable of performing legal research, document analysis, and other tasks designed to save time and increase efficiency. That is a good thing. Nevertheless, lawyers must consider the ethical challenges AI and generative AI present.

Generative AI is an artificial intelligence technology that produces new text, images, audio, and other content based on data. Generative AI represents a major advancement in artificial intelligence, enabling the creation of unique content beyond AI’s original scope. Generative AI handles tasks like document summarization and text editing, allowing others to manage other work more effectively. There are benefits to AI and generative AI (hereafter collectively, “AI”), but you cannot ignore the risks presented by its use, which includes, but are not limited to:

• Producing “hallucinations,” or inaccurate or incomplete output;
• Exposing confidential firm or client information to unauthorized persons;
• Exposing the firm or its lawyers to lawsuits alleging malpractice, breach of fiduciary duty, or other claims; and
• Violating ethical rules.

III. THE APPLICABLE TEXAS ETHICAL RULES

In Texas, the use of AI triggers a number of ethical duties. Accordingly, lawyers using AI have the ethical responsibility to understand the technology and how it works, the benefits and the risks of the technology, the obligation to verify all citations and materials in any filings, and to educate the client about your intention to use AI. In some instances, you may even need to obtain the client’s consent to the use of AI, for example, where confidentiality is a concern. Failure to satisfy any of these obligations runs the risk of violating various Texas Disciplinary Rules of Professional Conduct, including:

• Rule 1.01(a). Comment 8. Competence and Diligent Representation: A lawyer shall not accept or continue employment in a legal matter which the lawyer knows or should know is beyond the lawyer’s competence. Because of the vital role of lawyers in the legal process, each lawyer should strive to become and remain proficient and competent in the practice of law, including the benefits and risks associated with relevant technology;
• Rule 1.05. Safeguarding Client Information: Lawyers must safeguard information relating to the representation of a client and ensure that AI systems handling confidential data adhere to confidentiality obligations;
• Rule 1.09. Conflict of Interest: Lawyers must identify and address potential conflicts of interest arising from using AI;
• Rule 3.03. Candor Toward the Tribunal: Lawyers must ensure the accuracy and relevance of the citations provided by AI that one uses in legal documents or arguments;
• Rule 5.01. Responsibilities of a Partner or Supervisory Lawyer: Lawyers must supervise those using AI to ensure compliance with Texas Disciplinary Rules. A lawyer is subject to discipline because of another lawyer’s violation of the rules of professional conduct, if the lawyer is a supervising lawyer and encourages or knowingly permits the impermissible conduct involved; and
• Rule 5.05(b). Unauthorized Practice of Law: Lawyers must ensure that they are not assisting someone in the unauthorized practice of law. To the extent an AI programmer is not a lawyer and is providing legal analyses, that programmer may violate Rule 5.05 regarding the unauthorized practice of law.

IV. ABA FORMAL OPINION

On July 29, 2024, the American Bar Association (“ABA”) issued a formal opinion titled “Generative Artificial Intelligence Tools.” The opinion addresses the ethical ramifications of lawyers using AI in their practice and provides helpful guidance to practitioners.

Similar to Texas Disciplinary Rule 1.01(2), the ABA opinion reminds lawyers that they are required to have a reasonable understanding of the capabilities, limitations, benefits, and risks of using AI. In other words, because of the risk of inaccurate output or so-called “hallucinations,” independent verification of the output is necessary. Moreover, according to the opinion, users must evaluate the AI technology being used, analyze the output, not rely solely on the technology’s conclusions, and understand that AI cannot replace the attorney’s judgment.

The opinion also points out that lawyers are required to obtain the client’s informed consent before using AI. The lawyer therefore must inform the client of the use of AI, the risk of using it, and obtain the client’s informed consent for certain purposes before its actual use. Importantly, the opinion states that “merely adding general, boiler-plate provisions to engagement letters…is not sufficient” to meet this requirement.

V. OTHER STATES’ AI ETHICS DECISIONS

At least 16 states or courts have issued reports, ethics opinions, or judicial decisions focused on the ethical considerations inherent in the use of AI. For example, a Florida State Bar ethics opinion concluded that lawyers may use AI in the practice of law but must (1) protect the confidentiality of client information, (2) provide accurate and competent services, (3) avoid improper billing practices, and (4) comply with applicable restrictions on lawyer advertising.

Similarly, the New York State Bar Association Task Force Report provides (1) an extensive history and analysis of the evolution of AI, (2) the benefits and risks of its use, (3) the impact of AI on the legal profession, (4) a legislative overview and recommendations, and (5) AI guidelines under its Rules of Professional Conduct.

In California, the State Bar Standing Committee on Professional Responsibility and Conduct issued a report, explaining that:

“Generative AI use presents unique challenges; it uses large volumes of data, there are many competing AI models and products, and, even for those who create generative AI products, there is a lack of clarity as to how it works. In addition, generative AI poses the risk of encouraging greater reliance and trust on its outputs because of its purpose to generate responses and its ability to do so in a manner that projects confidence and effectively emulates human responses. A lawyer should consider these and other risks before using generative AI in providing legal services.”

The State Bar of Michigan addressed judicial competence and AI in an ethics opinion, concluding that judicial officers need to maintain competence with advancing technology, especially artificial intelligence, and how it affects their conduct and decisions. That ethics opinion provides examples of how AI can pose ethical dilemmas, such as bias, partiality, or accuracy, and how AI can assist judges in tasks like docket management, legal research, drafting documents, or answering questions.

Most recently, in 2024, the New Jersey Supreme Court Committee on Artificial Intelligence and the Courts issued its “Preliminary Guidelines On New Jersey Lawyers’ Use of Artificial Intelligence.” The guidelines explain that the use of AI does not change the lawyer’s duty to (1) be accurate and truthful, (2) be honest and candid when communicating, (3) preserve confidentiality, (4) prevent misconduct, including discrimination, and (5) provide oversight to lawyers, nonlawyer staff and others.

VI. TEXAS’ RESPONSE

In Texas, the State Bar’s Task Force for Responsible AI in the Law, created in 2023, has as its mission “to study the use of artificial intelligence in the legal profession, including ethical considerations, and make recommendations, if any, to the board of directors consistent with Tex. Govt. Section 81.012.” The task force issued an interim report, which “represents an initial step in understanding the integration of AI with the legal profession.” Among other things, the interim report warns that Texas lawyers using AI “face the risk of violating various disciplinary rules…”

Texas courts have been quick to address Texas lawyers’ use of AI. For example, Rule I (j) of the Local Rules of the Courts of Williamson County, Texas currently provides in pertinent part:

“While [AI] technology is developing quickly, it is currently unreliable and prone to bias and often fabricates information. The creators of these systems are not attorneys of record, licensed and in good standing to practice law in the State of Texas, and are not bound by the Texas Disciplinary Rules of Professional Conduct.

…

As a result, all self-represented litigants and attorneys who utilize any form of artificial intelligence for legal research or drafting in connection with a case shall, before using any AI-generated information in a court submission or proceedings, ensure that such information is accurate and correctly portrayed to the court, and shall sign and submit the Certificate Regarding Use of Artificial Intelligence (Exhibit A) and attach it to any filing where AI-generated information is used.”

Moreover, the Texas Supreme Court Advisory Committee and the Federal Rules Advisory Committee are now studying potential procedural and evidentiary rule changes to address this rapidly developing technology.

VII. COURTS’ RESPONSES TO AI

In addition to ethics opinions and the issuance of reports addressing the risks inherent in its use, courts also are beginning to address the problems associated with AI and are developing potential solutions. For example, some courts are mandating lawyer disclosures when any document submitted is based in whole or in part on AI. In 2023, a federal judge in the U.S. District Court for the Northern District of Texas entered a standing order requiring a Mandatory Certification Regarding Generative AI.14 According to the order, AI “is the product of programming devised by humans who did not have to swear [a lawyer’s] oath. As such, these systems hold no allegiance to any client, the rule of law, or the laws and Constitution of the United States.” Although the novel standing order made news, the judge has since removed that requirement from his court’s specific requirements. A fellow judge in the same district, however, has adopted essentially the same standing order, which provides:

“All attorneys and pro se litigants appearing before the Court must, together with their notice of appearance, file on the docket a certificate 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 checked for accuracy, using print reporters or traditional legal databases, by a human being.”

Similarly, a federal judge in Pennsylvania issued a standing order requiring counsel disclose whether they have used AI in the preparation of any complaint, answer, motion, brief, or other paper filed with the court, and a Magistrate of the U.S. District Court for the Northern District of Illinois issued a standing order requiring “[a]ny party using any generative AI tool in the preparation or drafting of documents for filing with the court must disclose in the filing that generative AI was used to conduct legal research and/or to draft the document.”

Other courts have entered similar rulings:

• A judge of the U.S. Court of International Trade issued an order on AI requiring disclosure of any generative AI program used and of all portions of text drafted with the assistance of generative AI, as well as certification that the use of the generative AI tool did not disclose confidential information to unauthorized parties. The judge expressed his concern that generative AI tools “that supply natural language answers to user prompts, such as ChatGPT or Google Bard, create novel risks to the security of confidential information.”
• A judge of the U.S. District Court for the Eastern District of Pennsylvania issued an order requiring the disclosure of any type of AI as opposed to limiting the disclosure requirement to the use of generative AI.
• A Magistrate of the U.S. District Court for the Northern District of California issued a standing order that addresses the difference between generative AI and tools that use other categories of AI, and stated that the disclosure requirement does not apply to the use of traditional AI, such as “the use of traditional legal research, word processing, spellchecking, grammar checking or formatting software tools (e.g., Lexis, Westlaw, Microsoft Word, or Adobe Acrobat).” The judge’s order addresses the disclosure requirements for AI and filings with the court, AI and confidentiality, AI and evidence which includes recognition that “AI-generated documents or materials (for example, created by a party prior to the commencement of litigation) are or may become exhibits, evidence or the subject of factual disputes in an action.”

VIII. SANCTIONS IMPOSED BY MISUSE OF AI

Misuse of AI has even resulted in sanctions being imposed on lawyers and their law firms. For example, in Mata v. Avianca, Inc., a court in the Southern District of New York sanctioned lawyers because the lawyers “abandoned their responsibilities” when they submitted an AI-written brief using ChatGPT and “then continued to stand by the fake opinions after judicial orders called their existence into question.” The individual lawyers and their law firm were fined $5,000 each. In People v. Crabill, a lawyer was suspended for one year and one day for citing cases created by ChatGPT that were not actual cases in a motion. The U.S. Court of Appeals for the Ninth Circuit recently struck a brief containing false authority drawn from AI.

Other courts have imposed various non-monetary sanctions attributable to the use of AI. The U.S. Court of Appeals for the Second Circuit referred a lawyer to a grievance panel for relying on ChatGPT without checking its results and for citing a non-existent decision in a reply brief. The Missouri Court of Appeals struck a pro se appellant’s appeal for filing a brief containing fictitious case citations generated by AI, reasoning it represented a flagrant violation of duties of candor owed to the court, which rose to the level of abuse of the judicial system.

IX. CONCLUSION

Given the risks involved with AI, both ethically and otherwise, lawyers should ask several questions before using AI in legal practice:

• Does the lawyer using AI possess sufficient knowledge of AI to satisfy the ethical duty of competence required by Texas Disciplinary Rule 1.01(2)? Has the client been informed of its use, and agreed?
• Who in the law firm will be responsible for supervising AI’s use, as required by Texas Disciplinary Rule 5.01?
• Is there a risk that using AI will result in confidential client information being communicated outside the firm, contrary to the requirements of Texas Disciplinary Rule 1.05?
• Does AI impermissibly use confidential client information provided on behalf of one client for any response to requests made on behalf of any other client, as prohibited by Texas Disciplinary Rule 1.09?
• What steps will be taken to ensure the accuracy of AI in any court or administrative filing, as required by Texas Disciplinary Rule 3.03?
• Will reliance on AI legal analysis provided by a non-lawyer run the risk that you are assisting in the unauthorized practice of law, according to Texas Disciplinary Rule 5.05(b)?

Courts, ethical committees, and lawyers are all struggling with how to utilize the new and developing AI technology, while at the same time honoring ethical obligations. Providing answers to these fundamental questions will go a long way toward satisfying those twin objectives.