New York has the most extensive AI regulatory framework for attorneys in the country. Between the NYSBA Task Force report, two NYC Bar formal opinions, a mandatory AI CLE requirement, and the Mata v. Avianca sanctions case, New York's 190,015 lawyers have more guidance, and more precedent, than any other state bar.
AI Regulation in New York: The Current Landscape
New York's AI regulatory landscape is unmatched in depth and breadth. The state has produced formal guidance from multiple bodies, created enforceable education requirements, and generated the single most cited AI sanctions case in American law. No other state combines all of these elements.
The NYSBA Task Force on Artificial Intelligence released a comprehensive report in April 2024, covering competence, confidentiality, and independent judgment when using AI. The NYSBA report is a task force report, not an ethics opinion. The New York City Bar has issued its own opinion on generative AI, which we did not audit, so this page does not number or describe it.
New York also implemented a mandatory CLE requirement: at least two annual credits in practical AI competency. This is enforcement through education, not just telling attorneys what to do, but requiring them to learn about it as a condition of maintaining their license. Combined with the state's 190,015 licensed attorneys and its status as the largest legal market in the country, New York's framework sets the national standard. New York's AI compliance layer extends beyond ethics opinions: attorneys advising employers must now account for NYC Local Law 144, which governs automated employment decision tools and creates client advisory obligations when AI is used in hiring, promotion, or termination decisions. On the federal court side, S.D.N.Y. judges have issued multiple individual AI disclosure standing orders that apply to any attorney practicing in the Southern District of New York, independent of the bar guidance.
What the New York Bar Says About AI
The NYSBA Task Force on AI Report (April 2024) provides the broadest framework. It addresses competence obligations, confidentiality requirements when using AI tools, and the duty of independent judgment: ensuring that lawyers don't defer to AI on matters requiring professional discretion. The report's recommendations are comprehensive and designed for a state bar that ranges from solo practitioners to the largest law firms in the world.
The NYSBA Task Force report suggests engagement-letter language on AI and says lawyers should seek assurance of a provider's confidentiality and monitor it. The New York City Bar has issued its own opinion on AI that we did not audit; read it on the NYC Bar site before relying on it.
Notably, New York does not require attorneys to inform or obtain consent from clients when using AI for general legal work. This distinguishes New York from California and Florida, which have client notification requirements. New York's position is that AI is a tool, like legal research databases or document management systems, and routine tool use doesn't require client consent. The consent requirement kicks in for specific situations, like recording client meetings.
Court Rules and Judicial Guidance
Mata v. Avianca, Inc. (S.D.N.Y. 2023) is the case that put AI sanctions on the national map. Judge P. Kevin Castel imposed $5,000 in sanctions on attorneys who submitted ChatGPT-generated fake case citations. The court found the attorneys acted with 'subjective bad faith' under FRCP Rule 11. The case was decided in the Southern District of New York (federal) but has been cited by courts nationwide as the benchmark for AI-related misconduct.
The Mata decision established several principles that apply beyond New York: attorneys are responsible for verifying every citation they submit, regardless of the source; using AI to generate legal research doesn't reduce the duty to confirm that cited cases actually exist; and courts will impose sanctions when attorneys fail to catch AI hallucinations that basic verification would have revealed.
Beyond Mata, New York courts are increasingly addressing AI in case management. The Southern and Eastern Districts of New York handle enormous federal caseloads, and judges in these courts have been among the most active nationally in issuing guidance on AI use in proceedings.
In late October 2026 the New York court system's Advisory Committee on Artificial Intelligence and the Courts published a seven-page guide for attorneys, "The Risks of Using Artificial Intelligence in Preparing Court Papers: What Every Attorney Should Know." We read the guide as hosted by News10 and News10's report on it. The court's own copy is not the one cited here, and the guide's text does not state its release day; news outlets reported it on October 26 and 27, 2026. The guide sorts court AI rules into four approaches (prohibition, certification, disclosure, informational) and places New York's statewide policy and model rule (22 NYCRR Part 161) in the informational group. It says the proposal does not require disclosure of AI use at filing and does not require a separate certification beyond the implicit one that comes from signing under 22 NYCRR § 130-1.1a, and that evidentiary materials are outside it. It tells attorneys to confirm every case, statute, regulation and quotation against an authoritative source such as Westlaw, Lexis, the official reporters or the court's docket, and says relying on the AI tool's own assurance is not verification. It also notes that individual judges' part rules may add obligations, so check the forum before filing.
Practical Implications for New York Attorneys
New York attorneys face the most comprehensive set of AI obligations in the country. The combination of NYSBA guidance, NYC Bar opinions, mandatory CLE, and Mata v. Avianca precedent means that every aspect of AI use in legal practice has been addressed. There's no excuse for non-compliance: the framework is complete.
The mandatory AI CLE requirement is the most direct impact on daily practice. Every New York attorney must complete at least two credits annually in practical AI competency. This isn't a recommendation: it's a licensure requirement. The credits must be in practical AI, not theoretical discussions, which means attorneys need to engage with actual tools and workflows.
For New York City firms, the competitive implications are significant. Clients choosing between firms look at AI capabilities as a differentiator. A firm that has integrated AI into its practice, trained its attorneys to meet the CLE requirement, and built governance frameworks aligned with the NYC Bar opinions is positioned differently than one that hasn't. In the most competitive legal market in the world, AI governance is now a client-facing issue.
What Attorneys in New York Should Do
Read the NYSBA Task Force Report and the New York City Bar's own opinion on generative AI. They address different aspects of AI in legal practice. We did not audit the NYC Bar opinion, so confirm its current text.
Get ahead of the CLE requirement. Don't treat the two annual AI credits as a checkbox: use them to build genuine competence with AI tools relevant to your practice area. Invest in CLE programs that cover hands-on use of generative AI, verification workflows, and confidentiality protocols. The requirement exists because the state determined that AI literacy is a professional necessity, not an optional skill.
If your firm uses AI meeting tools (transcription services, summary generators, recording platforms), get client consent before using AI to record or transcribe a client conversation, as a matter of Rule 1.4 and Rule 1.6 practice. Use platforms with security appropriate for privileged communications. Document your compliance.
The Bottom Line
New York has set the national standard for AI regulation in legal practice. Between formal opinions, mandatory CLE, task force guidance, and Mata v. Avianca sanctions precedent, there's no state where the expectations are clearer or the consequences more established. New York's 190,015 attorneys have everything they need: the only variable is whether they use it.
AI-Assisted Research. This piece was researched and written with AI assistance, reviewed and edited by Manu Ayala. For deeper takes and the perspective behind the research, follow me on LinkedIn or email me directly.
New: New York CLE includes 1 hour of cybersecurity, privacy and data protection each cycle. See AI CLE requirements and deadlines by state.