Likewise, the ABA Standing Committee on Ethics & Professional Responsibility has recently (July 29, 2024) weighed in with Formal Opinion 512, Generative Artificial Intelligence Tools (“Formal Op. 512”). This opinion highlights many of the same ethical rules as the other guidance, opinions, and reports, but identifies ethical requirements in a slightly different manner. Even though the approach of Formal Op. 512 is not binding on lawyers, it does offer a useful compilation of ethics guidance keyed to the Model Rules of Professional Conduct.
To be sure, GAI can significantly enhance efficiency in law practice because of the ability of this powerful tool to accomplish in a short timeframe a variety of tasks (e.g., legal research, document review, analysis of contracts and other legal documents) that used to be significantly more time-consuming. This, in turn, can benefit clients by reducing the often astronomical costs of legal services and possibly enhance access to justice by underserved populations.
Yet there are also significant pitfalls. Foremost among these are so-called “hallucinations”—a euphemism for nonexistent or false products of research, exemplified by notorious cases such as Mata v. Avianca, United States v. Michael Cohen, and the even more recent Second Circuit decision in Park v. Kim. Other common ethics hazards include potential breaches of client confidentiality and data privacy and succumbing to having one’s independent professional judgment swayed, if not entirely overborne, by the siren song of GAI work product.
Formal Op. 512 organizes the pertinent legal ethics implications of lawyers using GAI quite well. Here is a synopsis of the topics addressed:
- Competence (Model Rule 1.1): Lawyers must understand the capacity and limitations of GAI and periodically update that understanding.
- Communication (Model Rule 1.4): Formal Op. 512 offers guidance on when, and to what extent, lawyers are required to communicate their use of GAI to clients.
- Fees (Model Rule 1.5): Lawyers may not charge clients for time spent learning a technology to be used for client matters generally, unless a client specifically requests the use of a particular AI tool in a matter, in which case a lawyer may charge for learning how to use that particular tool. A lawyer may consider the cost of GAI to be part of office overhead, or may charge for a portion of an expensive and proprietary GAI tool, or on a per-use basis if appropriate, provided everything is fully explained to the client in advance and informed consent is obtained.
- Confidentiality (Model Rule 1.6): Lawyers are responsible for knowing how GAI uses data and putting in place adequate safeguards to ensure that data processed by GAI is secure and not susceptible to unwitting or unauthorized disclosure to third parties. Formal Op. 512 recommends that lawyers secure clients’ informed consent before using client confidences in GAI tools and opines that boilerplate consent included in engagement letters will not be adequate. The opinion also cautions that use by several lawyers of the same GAI tool may result in inadvertent use and disclosure of client information.
- Meritorious Claims and Contentions (Model Rule 3.1): Lawyers must be on their guard to prevent GAI “hallucinations” from forming the basis of asserting frivolous claims and arguments.
- Candor Toward the Tribunal (Model Rule 3.3 and Model Rule 8.4(c)): Overreliance on or uncritical adoption of GAI can result in making false statements of fact or law to a “tribunal” (which is broadly defined to include not only courts but others acting in an adjudicative capacity, potentially including arbitrators, administrative agencies, and legislative bodies). Duties to the tribunal require lawyers, before submitting materials, to review GAI output, including analysis and citations to authority, and to correct errors, including misstatements of law and fact, any failure to include controlling legal authority, and any misleading arguments.
- Responsibilities of Supervisory Lawyers (Model Rule 5.1): Partners and other lawyers with managerial duties must establish clear policies regarding the permissible use of GAI and supervise lawyer staff to ensure compliance with these policies. In this connection, Formal Op. 512 analogizes to principles in various state ethics opinions relating to cloud computing and outsourcing.
- Supervising Nonlawyers (Model Rule 5.3): Partners and other lawyers with supervisory responsibilities must make sure that nonlawyers (including those in a law firm as well as third-party contractors or other providers outside the law firm) are adequately trained in the ethical and practical uses of GAI.
Not discussed by Formal Op. 512 but worthy of consideration is another point:
- Bias and Prejudice in Law Practice and in Lateral Hiring (Model Rule 8.4(g)): As more firms routinely grow by lateral hiring, there is the potential for built-in bias or discrimination in any AI used to screen applicants and select from large numbers of résumés those to be interviewed. Lawyers must be careful to monitor the use of AI in hiring and results of such use.
In sum, GenAI, like a word processor or even a proprietary database like LexisNexis or Westlaw, is a tool, not a substitute for the exercise of legal expertise and judgment. Using GAI may streamline or even enhance lawyers in the performance of their tasks, but it does not absolve them of their responsibilities under applicable ethics rules.
