On March 13, 2026, California's State Bar Standing Committee on Professional Responsibility and Conduct (COPRAC) approved several proposed amendments to the state's Rules of Professional Conduct, specifically Rules 1.1 (Competence), 1.4 (Communication with Clients), 1.6 (Client Confidential Information), 3.3 (Candor), 5.1 (Responsibilities of Supervisory/Managerial Attorney), and 5.3 (Managing Non-Lawyers) ("the Rules") (the specific amendments can be seen on the State Bar website).
The proposed amendments primarily add comments to the Rules that address one central issue: the use of AI in tandem with legal research and writing, and the pitfalls associated with careless or inappropriate use of AI. Until now, the State Bar had not explicitly codified an attorney's duty with respect to the use of artificial intelligence into its Rules of Professional Conduct. The new amendments to the Rules accordingly create new grounds for attorney exposure and/or conduct that is sanctionable.
An important update to the Rulesis the addition of Comment 2 to Rule 1.1, "Competence." Comment 8 to Rule 1.1 previously stated, with emphasis: "To maintain [competency], a lawyer should keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology, engage in continuing study and education and comply with all continuing legal education requirements to which the lawyer is subject." The new comment language includes an explicit "duty to keep abreast of the changes in the law and its practice, including... artificial intelligence." Attorneys are now duty bound to "keep abreast" of changes with respect to the development of artificial intelligence. An attorney cannot avoid AI in 2026.
This duty not only applies to attorney work products, but also to attorneys who supervise non-attorneys per Rule 5.3, i.e., work product generated by a non-attorney. Proposed Comment 2 to Rule 5.3 states, with emphasis, that an attorney "must independently review, verify and exercise professional judgment regarding any output generated by the technology that is used in connection with representing a client." These sweeping changes largely address the use of AI in any capacity, and supervision of those who use AI, with respect to the performance of legal services. This is a much-needed change in the context of ongoing attorney discipline and malpractice exposure by virtue of an unchecked and unregulated use of AI.
The spirit of the proposed changes to the Rules concerns the widespread misuse of artificial intelligence by attorneys and law firms across the country, and the most frequent culprit is a reliance on false information presented by AI as fact, i.e., "hallucinations." Hallucinations can come in the form of completely fabricated caselaw and authority, i.e., a citation to a case that does not exist - or fabricated citations to a "real case," i.e., a scenario wherein the case exists, but the holding therein is fabricated.
Artificial intelligence in legal practice should not be seen as a point of exposure only. When harnessed appropriately, it can curb time spent and legal expenses for clients and can increase attorney efficiency. Thus, if - and when - one does choose to use artificial intelligence, one must treat the work product for what it is: AI-generated work product created by a non-lawyer. Consider this example. An extremely bright, intrepid and unusually fast-paced first-year associate completes a brief in a breathtakingly short amount of time. Her citations look right, and she is even citing well-known cases, in terms of the recognition of the specific cases being cited. No further review occurs, and the brief is then filed, resulting in the attorney filing a brief with hallucinated AI citations. Is the barely-cursory review appropriate for an attorney in terms of managing a non-attorney, in the context of presenting authorities to a court of law? Of course not. Attorneys must review each citation to ensure its accuracy. Treat AI like a junior law clerk, where oversight and guidance is very much still required.
There are new, preventative tools to help attorneys and firms navigate the increased exposure to attorney malpractice claims and/or the risks of using artificial intelligence in practice. Legal databases have implemented review tools designed to cross-reference citations that are used across all briefings to determine if a citation is "real" or not within a matter of seconds. This is a fantastic use of AI in the legal space. If attorneys are going to utilize a specific program for this purpose, there must be other safeguards in place. At an absolute minimum, there is an obligation to check and cross-reference every word, citation, or reference relied upon by an attorney and/or a non-attorney supervised by an attorney with respect to legal services. Otherwise, the attorney risks significant legal exposure.