Artificial Intelligence, or “AI,” has begun getting attention in the legal profession— particularly as it relates to ethical considerations and the Rules of Professional Responsibility. Many of the novel issues created by AI have not yet been considered by courts, however, nor are there any State laws or formal rules relating to AI that attorneys can rely on in practice.
In July 2024, the American Bar Association Standing Committee on Ethics and Professional Responsibility (the “Committee”) published Formal Opinion 512, which addresses the various effects that AI may have on the Model Rules of Professional Responsibility (the “Rules”). While not a single model rule even makes mention of AI, the Committee opined that attorneys should consider several different Rules prior to, and while using, such technology. This article provides a short summary of the Rules cited by Formal Opinion 512—as well as a short summary of the Committee’s Opinion as it relates to each Rule.
I. Rule 1.1 – Competency
Rule 1.1 requires attorneys to provide competent representation to their clients, and to have the requisite legal knowledge, skill and thoroughness in their representation. Comment 6 to Rule 1.1, which was the focus of the Committee’s Opinion, sets forth that lawyers need to remain cognizant of changes in the law and its practice, including the risks and benefits associated with the technology relevant to the attorney’s practice. While there is no requirement to use AI in practice to be considered “competent,” the Committee opined that if an attorney does use AI, he or she should understand the risks and benefits of the particular AI tool being used. Further, the attorney should understand the potential drawbacks in the technology in order to ensure that he or she does not give false or inaccurate information or advice to a client or tribunal.
II. Rule 1.6 – Confidentiality
Rule 1.6 requires attorneys to take reasonable steps to prevent the disclosure of confidential information—and there is no exception to the Rule for any kind of technology, including artificial intelligence that an attorney may use. The main consideration for this Rule in connection with AI is whether the technology an attorney is employing provides adequate assurance that the information is being used, stored and accessed in such a way that protects client confidentiality. For example, if the technology is “self-learning,” then there is a question as to whether the client information inputted by “Attorney A’s” firm may be used to return a response for “Attorney B’s” firm—thereby causing a question as to whether the Attorney A breached confidentiality. Attorneys should be sure to understand the manner in which any AI tool they take advantage of accesses and uses any confidential information from that practice.
III. Rule 1.4 – Communication
Rule 1.4 requires attorneys to inform clients of any and all decisions or circumstances needing the clients’ consent/approval. It also requires attorneys to consult with clients regarding the means which the attorney uses to accomplish the client’s objective. The Committee suggests that this Rule may require a client to be informed when it is likely that an attorney will use AI technology in working on their file–with the facts of each case determining whether a client needs to give consent. While it is currently unclear as to whether the client needs to give actual consent in every circumstance that a lawyer uses AI, it may be advisable to err on the side of caution and inform the client of the type of AI that may be used in a representation, as well as its purpose. This could potentially be accomplished through an engagement letter or retainer agreement where there is a paragraph dedicated to the potential use of AI. Through the client countersigning the document, he or she can give explicit consent to the attorney’s use of the tool.
IV. Rule 1.5 – Fees
Rule 1.5 is based on a standard of reasonableness: fees must simply be “reasonable.” Ultimately, this means that an attorney may only bill for the time actually spent on a file—not the time he or she could have spent but for the use of AI. In the context of AI, therefore, attorneys should have the understanding that while the assistance of AI may be beneficial in allowing them to reduce the time required on a particular file, it will, in turn, lower their billable hours. An attorney can still pass the cost of the technology to the client as an expense—as long as the client has given prior consent for the attorney to do so. But, absent advanced consent, the attorney may only charge for the direct cost associated with the technology. If the attorney wishes to add an additional surcharge above the direct cost, this would require additional informed consent by the client.
V. Rules 3.1, 3.3 and 8.4(c) – Meritorious Claims and Contentions and Candor Toward the Tribunal
This group of Rules, in the context of AI technology, requires attorneys to review any outputs prior to presentation to a court. While relatively self-explanatory on their face, the Rules provide that attorneys shall not engage in conduct involving dishonesty, fraud, deceit or misrepresentation. The Committee explicitly states in its Opinion that even an unintentional misstatement to a court can involve a misrepresentation under the Rules. Thus, an attorney should always check all output from AI to ensure there are no false statements or fictitious information contained therein.
VI. Rule 5.1 – Supervision of Subordinate Lawyers; and Rule 5.3 – Supervision of Non-Lawyers
These two Rules require attorneys to adequately supervise and train subordinate lawyers and non-lawyer employees. The Rules carry two considerations in the context of AI. First, managerial attorneys need to ensure that their subordinates are using AI technology that has been vetted and approved by the firm as one that conforms with the Rules. This could presumably be accomplished through an employment manual policy, as well as “adequate supervision” of employees. Second, if subordinates have permission to use certain AI technologies, managerial attorneys must ensure that the work being produced by the subordinates meets the requirements of the Rules. Ultimately, attorneys should see AI as an “artificial assistant”—and should supervise its work in the same manner in which they would supervise the work of their subordinate attorneys or legal assistants.