AI Competence Under NY Rule 1.1: A Training Framework for Law Firms

The Competence Gap

Ninety-five percent of legal leaders are using or implementing AI-enabled software.¹ Only ten percent of law firms have formal AI governance policies. That gap is not just a management failure — it is an ethics problem.

New York Rule of Professional Conduct 1.1 requires lawyers to “provide competent representation to a client,” which includes “the legal knowledge, skill, thoroughness and preparation reasonably necessary for the representation.” Comment 8 to ABA Model Rule 1.1 — adopted by an increasing number of jurisdictions — makes the technology dimension explicit: a lawyer should “keep abreast of changes in the law and its practice, including the benefits and risks associated with relevant technology.”

When the technology in question is generative artificial intelligence — capable of drafting briefs, summarizing depositions, reviewing contracts, and conducting legal research in seconds — the competence obligation takes on new urgency. Lawyers who use these tools without understanding them risk violating Rule 1.1. Lawyers who refuse to learn about them may face a different competence question: whether willful ignorance of available technology constitutes a failure to provide the “thoroughness and preparation reasonably necessary” for modern legal practice.

This article provides a practical framework for mid-size New York law firms — those with five to seventy-five attorneys — to build an AI training program that satisfies Rule 1.1 and the growing body of AI-specific ethics guidance.

The Regulatory Landscape in New York

Three overlapping sources of authority now define what competent AI use looks like for New York lawyers.

NYC Bar Formal Opinion 2024-5

Issued in August 2024, this opinion identifies the ethical obligations of lawyers and law firms relating to generative AI.² It maps existing Rules of Professional Conduct onto AI use, identifying duties across confidentiality, competence, supervision, candor, client communication, and fees. Its core message: lawyers must understand “to a reasonable degree how the technology works, its limitations, and the applicable Terms of Use and other policies governing the use and exploitation of client data by the product.”

The opinion explicitly requires that lawyers not rely on AI-generated information without independent verification. It further distinguishes between “open” AI systems that may share data with third parties and “closed” enterprise systems — a distinction with direct implications for confidentiality obligations under Rule 1.6.

Notably, the NYC Bar adopted a “guardrails, not hard-and-fast restrictions” approach, recognizing that rigid rules would be outdated before the ink dried. This makes competence-through-training, rather than compliance-through-prohibition, the operative framework.

NYSBA Task Force on Artificial Intelligence

The NYSBA Task Force released its Report and Recommendations in April 2024, representing the most comprehensive guidance from a state bar association at the time.³ Its four principal recommendations prioritize education over legislation:

1. Adopt guidelines for attorney use of AI and generative AI
2. Prioritize education of judges, lawyers, law students, and regulators
3. Identify risks not addressed by existing law for potential new regulation
4. Examine the function of law as a governance tool for AI

The Task Force’s emphasis on education is particularly relevant for firm leadership. The report envisions a profession that learns to apply existing ethical frameworks to new technology — not one that waits for regulators to issue specific AI rules. For managing partners, this means the obligation to train attorneys on AI use exists now, under current rules, regardless of whether specific AI regulations follow.

ABA Formal Opinion 512

Released in July 2024, ABA Formal Opinion 512 was the national profession’s first comprehensive ethics guidance on generative AI.⁴ It addresses six areas: competence (Model Rule 1.1), confidentiality (Model Rule 1.6), communication with clients (Model Rule 1.4), candor toward the tribunal (Model Rules 3.1 and 3.3), supervisory responsibilities (Model Rules 5.1 and 5.3), and reasonable fees.

Two points from Opinion 512 deserve special attention from firm leadership:

Supervision is an affirmative obligation. Under Rules 5.1 and 5.3, partners with managerial authority must establish firm-wide policies on AI use and ensure that both lawyers and non-lawyers are adequately supervised in their use of AI tools. This is not aspirational language — it is an existing ethical requirement that simply had no AI-specific application until now.

Billing must reflect reality. If an AI tool reduces a ten-hour contract review to two hours of attorney time, the firm cannot bill for ten. Opinion 512 also clarifies that lawyers generally may not charge clients for time spent learning how to use AI tools, unless the client specifically requested the use of a particular tool.

Court-Level Requirements

Beyond ethics opinions, individual judges and court systems have created enforceable AI requirements. The NY Unified Court System mandated AI training for all court personnel effective October 2025. While this directive applies to court staff rather than attorneys, it signals the judiciary’s expectation that all participants in the legal system — including attorneys appearing before the court — understand AI.

Multiple judges in the Southern and Eastern Districts of New York have issued standing orders or individual rules requiring:

• disclosure when AI tools are used in preparing court filings
• certification that AI-generated content has been independently verified
• confirmation that all filings comply with the requirements of Rule 11

A proposed New York State Senate bill would formalize these requirements, mandating that any attorney who used generative AI to draft legal documents disclose that use upon filing and certify that a human reviewed the submission.⁵

For firms litigating in federal courts, the practical takeaway is clear: attorneys need to understand their AI disclosure obligations on a judge-by-judge basis, and firm policy must account for this patchwork of requirements.

A Three-Phase Training Framework

The following framework is designed for firms that need to move from ad hoc AI use to a defensible, documented training program. It is calibrated for firms with five to seventy-five attorneys and limited internal technology resources.

Common Failure Modes

Firms that have attempted AI adoption without a structured training program tend to fail in predictable ways:

The “unofficial tool” problem. Associates use ChatGPT or similar tools on their personal devices, outside the firm’s security environment, without supervision or policy guidance. The firm has no visibility, no control, and no defensible position if something goes wrong.

The “one-and-done” training. A single CLE session on “AI and the Law” does not satisfy the ongoing supervisory obligations of Rule 5.1. Competence is not a box to check — it is a continuous duty that requires continuous training.

The “policy without training” mistake. Some firms have drafted AI policies without providing the training necessary to comply with them. A policy that prohibits “entering confidential client information into unapproved AI tools” is useless if attorneys do not understand which tools are approved, which are not, and why.

The “ban everything” overcorrection. A blanket prohibition on AI use does not satisfy Rule 1.1’s competence requirement. If AI tools offer material benefits to clients — faster research, lower costs, more thorough review — a firm that prohibits their use may face questions about whether it is providing the “thoroughness and preparation reasonably necessary for the representation.”

Conclusion

The duty of competence has always required lawyers to keep pace with the tools of the profession. When typewriters gave way to word processors, when Westlaw replaced the library stacks, when email displaced the postal service, the profession adapted — not because regulators mandated it, but because competent practice demanded it.

Generative AI is the same kind of inflection point, but faster. The ethics opinions are clear. The court requirements are multiplying. The enforcement actions — from Mata v. Avianca to the growing number of sanctions for unverified AI-generated citations — provide cautionary examples.

For managing partners and firm leadership, the path forward is straightforward: assess your firm’s current AI use, draft a policy that meets your ethical obligations, train every attorney and staff member against that policy, and maintain the program as the rules evolve. This is not an innovation initiative. It is a compliance requirement under rules that already exist.

The firms that build this infrastructure now — while the regulatory landscape is still forming — will be better positioned than those that wait for a crisis to force the issue. The NYSBA Task Force had it right: education first, legislation later. The time for that education is now.

---

---

---

Tyler Brennan is a New York-licensed attorney specializing in legal ethics and technology compliance. Ansgar Lange is co-founder of Fractal Legal, an AI training and policy subscription service for New York law firms. The views expressed in this article are the authors’ own.