Part of: Court AI Standing Orders by Jurisdiction
- 22 NYCRR Part 161, “Use of Artificial Intelligence Technology,” was adopted March 25, 2026 and took effect June 1, 2026, covering every court in the New York Unified Court System, civil and criminal (New York Courts, Part 161).
- The system-wide default under Part 161 is that attorneys and parties are not required to disclose AI use to the court — the rule states plainly that “attorneys and parties should not be required to disclose to the court that they have used AI tools,” since existing professional duties already apply regardless of how a filing was prepared.
- Individual courts may opt into a stricter Model Rule set out in Appendix A, under which a filer’s signature certifies the paper contains no fabricated or fictitious content — but that certification obligation applies only in courts that affirmatively adopt it, not statewide by default.
- The rule defines “paper” as a brief, memorandum, affidavit, affirmation, pleading, or other document prepared for submission to a court — explicitly excluding materials constituting or offered as evidence, which are handled under separate evidentiary rules.
What Part 161 actually says, read directly
New York’s approach is easy to summarize incorrectly, because “New York has a statewide AI court rule” sounds like it should mean a uniform disclosure requirement, and it doesn’t. The rule’s core substantive requirement is this: “Any attorney or party who uses an artificial intelligence tool in preparing any paper filed in or submitted to a court is required to carefully review the paper and independently ensure that it contains no fabricated or fictitious cases, statutes, or other material. By signing such paper, an attorney or party certifies that such a review has been conducted and that the paper contains no such fabricated or fictitious content” (New York Courts, Part 161). That obligation exists system-wide, effective June 1, 2026.
But the rule’s policy section is explicit that this obligation does not translate into a disclosure requirement: “the use by attorneys and parties of artificial intelligence tools in preparing papers submitted to a court should not be prohibited, as long as such use is in accordance with the duties and responsibilities that apply to individuals who submit papers to a court. Since those duties and responsibilities already apply to all submissions, attorneys and parties should not be required to disclose to the court that they have used AI tools.” In other words: the underlying accuracy obligation is universal and mandatory across every New York Unified Court System court; the requirement to affirmatively tell the court AI was used is not.
The opt-in Model Rule, and why “statewide rule” is misleading
Part 161 gives individual courts discretion to go further: “A court may, in its discretion, implement a part rule governing the use by attorneys and parties of artificial intelligence tools in preparing papers submitted to the court. If a court determines that such a rule is appropriate, the court is encouraged to adopt the model rule set forth in Appendix A of this Part.” The Model Rule, where a specific court adopts it, is the mechanism that produces the certification-and-disclosure structure more commonly associated with AI standing orders elsewhere in this cluster.
This two-tier structure means the practically important question for any filer is not “does New York have an AI rule” — it does, and it applies everywhere in the system — but “has the specific court I’m filing in adopted the Appendix A Model Rule.” Two litigants filing in different New York courts on the same day, using AI tools identically, could face different disclosure obligations depending entirely on whether their particular court opted into the stricter tier. A jurisdiction-by-jurisdiction check of which specific New York courts have adopted the Model Rule is necessary before assuming either the default (no disclosure) or the stricter tier (certification) applies to a given filing.
What “paper” does and doesn’t cover
The rule’s scope is defined narrowly and specifically: “paper” means “a brief, memorandum, affidavit, affirmation, pleading, or other document prepared by an attorney or party for submission to a court,” and this definition explicitly excludes “materials constituting or proffered as evidence in the case, as such materials are subject to separate considerations and requirements” (New York Courts, Part 161). This carve-out matters because it means Part 161’s certification obligation, where applicable, attaches to argumentative and procedural documents — the kinds of filings where a fabricated citation would actually appear — not to evidentiary submissions, which are governed by New York’s separate evidence rules (including any AI-specific authentication requirements those rules may impose, which sit outside Part 161’s scope).
Why this design choice makes sense on its own terms
Part 161’s system-wide default of “the underlying obligation is universal, disclosure is not” reflects a specific policy judgment worth stating plainly: the rule’s drafters concluded that requiring affirmative AI-use disclosure adds a compliance step without necessarily improving accuracy, since the accuracy obligation (no fabricated content) already binds every filer regardless of method. A certification-model rule like Judge Wang’s Colorado order (see the companion spoke on the Starr standing order) essentially agrees with this logic but goes one step further by making the certification itself the enforcement mechanism — Part 161’s default leaves enforcement to standard sanctions processes (grounded in existing rules of professional conduct and court authority) rather than creating a new procedural gate, unless a specific court opts in via Appendix A.
What this means for anyone filing in a New York court
- Check whether your specific court has adopted the Appendix A Model Rule before assuming either tier applies. The statewide default is no mandatory disclosure; a specific court’s adoption of the stricter tier changes that for filings in that court only.
- The no-fabrication obligation applies everywhere in the system regardless of the Model Rule question. Even in courts that haven’t opted into Appendix A, Part 161’s core requirement — carefully reviewing AI-assisted content and ensuring no fabricated material — is already binding as of June 1, 2026.
- Evidence is handled separately. Part 161’s “paper” definition excludes evidentiary materials; don’t assume this rule covers AI-generated or AI-authenticated evidence, which is governed by New York’s separate evidence framework.
- This is a new rule with limited case law behind it. Effective June 1, 2026, Part 161 has not yet accumulated the body of enforcement decisions that older rules like Judge Starr’s have — treat guidance interpreting its practical application, including this article, as provisional pending actual court decisions applying it.
Related reading
- Court AI Standing Orders by Jurisdiction — the pillar this spoke expands on.
- Judge Brantley Starr’s AI Standing Order: The Template Every Court Copied — the certification-model comparison point.
- citationsafe.com/sanctions-database
Sources
- New York State Unified Court System, 22 NYCRR Part 161, Use of Artificial Intelligence Technology — official rule text, quoted directly (core requirement, disclosure policy, Model Rule provision, and “paper” definition), effective June 1, 2026.
- New York State Bar Association, “Effective June 1, 2026, The New York State Unified Court System Has Adopted a New Rule Regarding the Use of Artificial Intelligence”, corroborating summary.
Citation Safe is a verification workflow tool, not legal advice. No attorney-client relationship is created. Human review is always required. Court rules change; always confirm current requirements — including whether a specific court has adopted the Appendix A Model Rule — against the New York court system’s own published rules before filing.