Late last month, the American Bar Association issued Formal Opinion 512 detailing a lawyer’s ethical obligations when using generative artificial intelligence (genAI) in legal practice.

The Opinion acknowledges that lawyers often use traditional AI to boost the quality and efficiency of their legal practices, for example in contract analysis or e-discovery. However, the advent of genAI, which creates new content based on user prompts or questions, raises unique issues related to the ABA Model Rules of Professional Conduct.

The Opinion offers general guidance for lawyers in six specific areas but notes that given the rapid evolution of genAI tools, updated guidance will likely be required as the technology develops.

Below, we outline the main points from the six areas in which the ABA offers guidance.

Competence

Lawyers do not need to become genAI experts to competently use genAI tools in legal representation, the Opinion says, but they “must have a reasonable understanding of the capabilities and limitations” of the tools they might use. This includes knowing the risks.

Lawyers can develop an understanding of genAI tools through reading, CLEs, or other programs, or by drawing on the knowledge of experts. Lawyers must view this education as ongoing, as the capabilities of genAI tools are constantly evolving.

In addition, given genAI’s potential to produce false and inaccurate information, lawyers must independently verify any genAI outputs that are used in client representation, the Opinion says.

Even if a lawyer does not expect to use genAI when representing clients, the Opinion notes that “lawyers should become aware of the [genAI] tools relevant to their work so that they can make an informed decision, as a matter of professional judgment, whether to avail themselves of these tools or to conduct their work by other means.”

Confidentiality

The Opinion points to a lawyer’s duty to confidentiality of client information under Model Rule 1.6 as well as their duty to “make reasonable efforts to prevent the inadvertent or unauthorized disclosure of, or unauthorized access to, information relating to the representation of a client.”

This obligation also applies to former and prospective clients, as outlined in Model Rules 1.9(c) and 1.18(b).

Therefore, the Opinion states that before lawyers enter information related to client representation into a genAI tool, they must assess the potential that the information entered into the tool will be “disclosed to or accessed by” other individuals inside and outside the firm.

This includes the potential for disclosure of confidential client information in outputs generated by other individuals who use the same “self-learning” genAI tool. The Opinion describes self-learning genAI tools as those that learn from themselves as they process more data (e.g., ChatGPT).

Therefore, the Opinion states that informed client consent is required before entering information related to client representation into a self-learning genAI tool. And this informed consent must go beyond adding “boiler-plate provisions” to engagement letters, the Opinion says.

Communication

When disclosure and informed consent of the use of genAI is not required by Model Rule 1.6, the Opinion states lawyers must consider whether disclosure is required under other Model Rules, specifically Model Rule 1.4 on communications.

The Opinion identifies examples of when disclosure is required under Model Rule 1.4, for instance, regarding the reasonableness of lawyer’s fees or if the lawyer is using genAI to inform important decisions about the representation.

No matter the situation, in the spirit of effective communication, it may be prudent for lawyers to explain how they use genAI generally in the delivery of legal services a client engagement letter, along with any instructions for clients, the Opinion says.

Meritorious Claims and Contentions and Candor Toward the Tribunal

The Opinion identifies lawyers’ ethical obligations to the courts under Model Rules 3.1, 3.3, and 8.4(c), primarily centered on presenting factual information and assertions that do not misrepresent the facts in a case.

However, given genAI’s potential to produce inaccurate information, the Opinion reinforces the necessity for lawyers to ensure the accuracy of all genAI outputs related to client representation, including before they are used in judicial proceedings.

Supervisory Responsibilities

Lawyers in managerial and supervisory roles have ethical obligations toward the firm, other lawyers, and other legal professionals associated with the lawyer under Model Rules 5.1 and 5.3. These obligations include these individuals’ use of genAI related to client representation, the Opinion states.

Lawyers in managerial and supervisory roles should “establish clear policies” on the use of genAI at the firm as well as training for subordinate lawyers and other legal professionals in the firm on genAI, including the ethical and practical uses and risks in client representation.

This supervisory responsibility also extends to lawyers who “rely on others outside the law firm” to use genAI related to client representation. For example, this responsibility may include cloud-computing platforms that store client data and the use of third-party service providers.

Fees

The Opinion guides that Model Rule 1.5 applies to lawyer fees when genAI is used during the engagement. This involves explaining charges related to the use of genAI before beginning the representation or at a reasonable time thereafter.

When determining the reasonableness of fees, lawyers using genAI should account for its potential to help lawyers complete tasks faster. They should also “analyze the characteristics” of each genAI tool and how it will be used in client representation before deciding if the tool is overhead or an out-of-pocket cost that can be billed to the client (see ABA Formal Opinion 93-379).

In addition, because lawyers are obligated to provide competent representation, they may not charge clients for time used to educate themselves about genAI tools that they will “regularly use for clients” in legal practice, the Opinion states.

It may, however, be appropriate to charge clients for the time used to learn a genAI tool that the client specifically requested be used in their representation, as long as the agreement is discussed beforehand.

State opinions on genAI in legal practice

In addition to the ABA’s Formal Opinion 512, some states and the D.C. Bar have provided guidance on the use of genAI in legal practice, including California, Florida, KentuckyMichigan, Missouri, New Jersey, and Pennsylvania.

In Illinois, “Generative AI and the Illinois Judicial Branch” is one of nine strategic initiatives the Illinois Judicial Branch identified in its 2024 Operational Plan.

As part of this, an AI Task Force created by the Illinois Judicial Conference (IJC) is exploring how to effectively use genAI tools to improve efficiency in the courts and court user experience, while considering policies needed to guide its use.

In addition, the Illinois State Bar Association (ISBA) launched a Standing Committee on Artificial Intelligence and the Practice of Law that is evaluating legislation, rules, and policies related to genAI and passing this education on to its members and the public.

“I remain both excited and optimistic about how generative AI can transform the legal profession,” Mark Palmer, Chief Counsel of the Illinois Supreme Court Commission on Professionalism who also sits on the ISBA Board of Governors and the IJC’s AI Task Force, told the Illinois Bar Journal.

“Just as other technologies have entered our space over the years—like computers, the internet, email, and cloud computing—we’re going to have to start with education,” he said. “The ISBA and other bar associations are well-positioned to foster awareness and education of responsible and ethical adoption of AI technologies in the legal profession. Bar associations can and should be an intermediary between legal tech companies developing novel tools and applications and the legal professionals using them.”

Palmer often writes on the use of genAI in legal practice for the Commission’s blog. Read more here.

Staying up to date on issues impacting the legal profession is vital to your success. Subscribe here to get the Commission’s weekly news delivered to your inbox.

The AI Revolution in Legal and the Billable Hour: Is the End Near?

3 Things Lawyers Should Be Doing Now with Generative AI

The Rise of ChatGPT: Ethical Considerations for Legal Professionals

The post Breaking Down the ABA’s Guidance on Using Generative AI in Legal Practice appeared first on 2Civility.