Comment Letter on OCA Generative AI Proposal

SUMMARY

The Presidential Task Force on Artificial Intelligence and Digital Technologies (Lorraine McGowen, Tiffany Smith, and Jerome Walker, Co-Chairs), Litigation Committee (Cassandra L. Porsch, Chair), and Professional Ethics Committee (Michael E. Salzman, Chair) submitted a letter in response to a request for comment from the New York State Office of Court Administration (OCA) on a proposal to add a new Part 161 to the Rules of the Chief Administrator of the Courts (RCAC) regarding the use of generative artificial intelligence (GenAI) in preparing court documents (GenAI Proposal). The committees fully support the GenAI Proposal and offer recommendations on how to further improve the proposal, including specific changes in the Model Rule in Appendix A.

REPORT

By Email

David Nocenti, Esq.
Office of Court Administration
25 Beaver Street, 10th Floor
New York, NY 10004
rulecomments@nycourts.gov

Re: Proposal to add a new Part 161 to the Rules of the Chief Administrator of the Courts regarding the use of generative artificial intelligence in preparing court documents

Dear Mr. Nocenti:

The New York City Bar Association[1] (City Bar) Presidential Task Force on Artificial Intelligence and Digital Technologies[2] (Task Force), Litigation Committee[3] (Litigation Committee), and Professional Ethics Committee[4] (Ethics Committee) submit this letter in response to a Memorandum dated November 17, 2025 from the New York State Office of Court Administration (OCA) soliciting public comment on a proposal to add a new Part 161 to the Rules of the Chief Administrator of the Courts (RCAC) regarding the use of generative artificial intelligence (GenAI) in preparing court documents (GenAI Proposal). The GenAI Proposal, prepared by the Advisory Committee on Artificial Intelligence and the Courts, consists of (i) a new Part 161, (ii) an Appendix thereto with a suggested Model Rule for judges who choose to implement a rule governing the use of GenAI by attorneys and parties, and (iii) a memorandum (Memorandum) in support of (i) and (ii).

The Task Force, the Litigation Committee, and the Ethics Committee appreciate the opportunity to comment[5] on the GenAI Proposal and the efforts of OCA in addressing in a thoughtful way an issue that is critically important to the Task Force, the Litigation Committee, the Ethics Committee, and stakeholders in New York and across the country. The Task Force, the Litigation Committee, and the Ethics Committee fully support the GenAI Proposal, though we make recommendations on how to further improve the GenAI Proposal. The Task Force, the Litigation Committee, and the Ethics Committee emphasize that GenAI capabilities are advancing at a rate that may render technology-specific disclosure rules outdated. What is today understood as GenAI may soon be supplemented or replaced by more complex agentic artificial intelligence systems capable of autonomous research, analysis, and drafting. History shows that lawyers have incorporated technological innovations, from electronic research databases to word-processing software, without the need for disclosure requirements tied to particular tools. As stated, existing ethical rules already obligate attorneys to ensure the accuracy of their submissions and to competently supervise any technologies they use.

Task Force, Litigation Committee, and Ethics Committee Research and Analysis Support the GenAI Proposal

The Task Force, the Litigation Committee, and the Ethics Committee have conducted extensive research and analysis of the issues underlying the GenAI Proposal over the course of many years, and many members of the Task Force, the Litigation Committee, and the Ethics Committee have researched and analyzed the underlying issues for even a longer period of time. Indeed, in August 2023, in response to Mata v. Avianca, Inc., the Task Force released a podcast[6] addressing the implications of using GenAI in connection with court filings in New York. Similarly, in November 2023, the Task Force released another podcast[7] in response to United States v Michel and the increasing judicial requirement for lawyers who file court documents and use GenAI in connection with those court documents to disclose the use of GenAI. The podcast was necessary because the Task Force became increasingly concerned that requiring disclosure of GenAI use was unwarranted given that courts already had sufficient remedies to address the underlying issues that were surfacing as a result of the use of GenAI in court filings. In response to more court action involving disclosures, the Task Force released another podcast[8] in April 2024. This time the podcast included two Task Force members who are former U.S. District Court judges.[9]

By January 2025, there was even greater concern within the Task Force that an ever-increasing use of court orders and other rulings might themselves be adding to the confusion around the use of GenAI in court filings. The Task Force released yet another podcast[10] on the subject matter, focused entirely on the issue of whether the disclosure of the use of GenAI should be required by lawyers and judges. This podcast included two Task Force members, a sitting U.S. District Court judge and a Computer Science Professor,[11] who have made several presentations to bar associations and judicial groups on the subject matter.

Simultaneously with the research and analysis related to the podcasts, in June 2024, the Task Force and the City Bar Council on Judicial Administration[12] issued a report[13] that, among other things, emphasized that (1) lawyers should not be required to disclose their use of AI in preparing court papers; (2) existing ethical rules were sufficient for governing the situation concerning lawyer use of AI; and (3) judges should consider limiting themselves in issuing individual rules regarding use of AI.  On this last point, the report expressly provides that:

For individual judges who may be considering issuing guidance [on the use of AI], the Working Group suggests that the court consider limiting such guidance to reminding attorneys that they must adhere to their ethical obligations when using generative AI, including having a working knowledge of how this technology functions and its attendant risks.

Many of the reasons used by OCA to explain its position mirror the same explanations used by the Task Force in its podcasts and in the report. Indeed, the three points identified above from the report are included as integral parts of the GenAI Proposal. For example, the supporting Memorandum indicates that:

The AI Advisory Committee notes that attorneys and parties are already required to ensure that papers they sign are reviewed for accuracy. If the attorney or party submits a paper containing false or fictitious matter, they are already subject to sanctions or discipline, regardless of whether AI technology was used in the preparation of the paper.

More specifically, as the Memorandum points out, state court decisions have already noted that existing RCAC § 130-1.1 authorizes sanctions against lawyers who fail to correct submissions with fictional citations.

While the Task Force, the Litigation Committee, and the Ethics Committee fully support the GenAI Proposal, the Task Force, the Litigation Committee, and the Ethics Committee also recommend a change to the following language in the Model Rule in Appendix A:

Accordingly, any attorney or party who uses a generative AI tool, as defined in 22 NYCRR 161.2(b), in preparing any paper, as defined in 22 NYCRR 161.2(c), filed in or submitted to this court or served on another party in a case before this court is required to carefully review the paper and independently ensure that it contains no fabricated, fictitious, or otherwise non-genuine cases, statutes, citations, quotations, or other material.

It is not unusual for a large language model to cite a correct case, but for the wrong proposition, or make an error such as mistaking a dissent for the main opinion. In addition, hallucinations sometimes take the form of asserting as a purported quotation words that are not contained in the cited source material, even if that source material itself that is cited is genuine and a reference to “citations” would help deal with source material not properly labeled (e.g., as a dissent or a concurrence, or wrong court/judge/year, etc.). Therefore, the Task Force, the Litigation Committee, and the Ethics Committee recommend that the above highlighted language be added, thus broadening the effective definition of hallucinations.

In addition, the Task Force, the Litigation Committee, and the Ethics Committee recommend a change to the following language in the Model Rule in Appendix A:

 (b) The term “generative AI” shall mean an AI program or system that is capable of generating text or other content in response to user prompts or other inputs by being trained on material in large reference datasets.

As technology advances, it is likely that AI programs and systems will be capable of generating text or other content in the absence of a user’s prompt.

The Memorandum also recognizes that individual judges may nevertheless find it useful to adopt a rule concerning the use of GenAI in order “to alert attorneys and unrepresented parties to the risks associated with the use of generative AI, and to ensure that they are on notice as to the possible consequences of misusing it.”  The Task Force, the Litigation Committee, and the Ethics Committee believe it is sensible that the proposed Part 161 includes a model rule that judges across the state can adopt, thereby promoting uniformity as well as avoiding the burden for each judge to “reinvent the wheel” with his/her own GenAI rule.

Conclusion

The findings of multiple years of research and analysis by the Task Force, the Litigation Committee, and the Ethics Committee are consistent with the GenAI Proposal, and the Task Force, the Litigation Committee, and the Ethics Committee agree with OCA that it is not advisable for a court to impose an outright ban on the use of any GenAI technology. Accordingly, and subject to the recommendation above, the Task Force, the Litigation Committee, and the Ethics Committee fully support the GenAI Proposal, and we urge OCA to adopt the slightly modified proposal.

Respectfully,

Presidential Task Force on Artificial Intelligence and Digital Technologies
Lorraine McGowen, Co-Chair
Tiffany Smith, Co-Chair
Jerome Walker, Co-Chair

Litigation Committee
Cassandra L. Porsch, Chair

Professional Ethics Committee
Michael E. Salzman, Chair

Task Force Subcommittee on the Use of Artificial Intelligence in the Judiciary
Harut Minasian, Co-Chair
David Zaslowsky, Co-Chair