ABA Formal Opinion 512: A Compliance Guide

What Opinion 512 is, and why it matters

Formal Opinion 512 is the ABA Standing Committee on Ethics and Professional Responsibility's first formal opinion on generative artificial intelligence. Issued July 29, 2024, the Opinion applies six Model Rules to lawyer use of GAI. The obligations are framed as follows:

"To ensure clients are protected, lawyers using generative artificial intelligence tools must fully consider their applicable ethical obligations, including their duties to provide competent legal representation, to protect client information, to communicate with clients, to supervise their employees and agents, to advance only meritorious claims and contentions, to ensure candor toward the tribunal, and to charge reasonable fees."
Opinion 512, syllabus (p. 1).

The opinion is advisory, not binding: the ABA Model Rules are not the rules of professional conduct in any jurisdiction until a state supreme court adopts them. Opinion 512 has become the de facto national baseline for two reasons. First, 49 of 50 states (California is the outlier) have adopted the Model Rules' core structure, so the Opinion's rule-by-rule framing maps cleanly onto most state RPCs. Second, state bar AI opinions issued since July 2024 routinely cite and track Opinion 512's analysis, with state-specific overlays rather than wholesale departures.

Practical consequences are concrete. If a firm cannot document that its AI practices address each of the six rules below, a malpractice carrier, a disciplinary investigator, or a sanctioning court has a ready-made framework for identifying gaps. Each section below translates a rule into the obligation and the documentation firms should keep.

Rule 1.1: Competence

Model Rule 1.1 requires competent representation. Comment [8] extends that duty to the benefits and risks of relevant technology. Opinion 512 applies this to GAI by stating the standard plainly:

"To competently use a GAI tool in a client representation, lawyers need not become GAI experts. Rather, lawyers must have a reasonable understanding of the capabilities and limitations of the specific GAI technology that the lawyer might use."
Opinion 512, at 2–3.

The duty is ongoing. A one-time training does not satisfy Rule 1.1 years later if the tool, the model, or the firm's use of it has materially changed. The Opinion is explicit:

"This is not a static undertaking. Given the fast-paced evolution of GAI tools, technological competence presupposes that lawyers remain vigilant about the tools' benefits and risks."
Opinion 512, at 3.

Opinion 512 identifies acceptable ways to maintain competence: reading about GAI tools aimed at the legal profession, attending CLE, and consulting colleagues or external experts proficient with the specific tool in use. Practical documentation here is narrow. Keep records showing that attorneys and staff have been trained on each approved tool, and that training was refreshed when the tool materially changed.

Rule 1.6: Confidentiality

Rule 1.6 prohibits a lawyer from disclosing client information without informed consent. Rule 1.6(c) requires reasonable efforts to prevent inadvertent or unauthorized disclosure. Opinion 512 applies both duties to GAI input. Self-learning tools draw particular emphasis. Prompts and client data may be retained and surface later in another user's session.

When informed consent is required, the Opinion is explicit:

"[B]ecause many of today's self-learning GAI tools are designed so that their output could lead directly or indirectly to the disclosure of information relating to the representation of a client, a client's informed consent is required prior to inputting information relating to the representation into such a GAI tool."
Opinion 512, at 7.

Informed consent in this context is not the language most engagement letters already contain. The Opinion is direct about this:

"To obtain informed consent when using a GAI tool, merely adding general, boiler-plate provisions to engagement letters purporting to authorize the lawyer to use GAI is not sufficient."
Opinion 512, at 7.

To satisfy Rule 1.6, the Opinion describes a baseline review duty. Every lawyer using a GAI tool should read the Terms of Use, privacy policy, and related contractual terms for that tool. The review may be delegated to a colleague or external expert. Four questions must be answered before inputting client information:

• Who has access to inputs and outputs.
• Whether and how the provider retains data.
• Whether the tool trains on submitted content.
• How the tool behaves if access is revoked.

Document the vendor review for each approved tool. Keep a record of which matters, if any, required tool-specific informed consent.

Rule 1.4: Communication with clients

Communication with clients about the means of representation falls under Rule 1.4. Opinion 512 imposes no blanket duty to disclose GAI use on every matter. Facts control. Disclosure is required in five categories:

• The client asks whether GAI was used.
• The engagement letter or outside counsel guidelines require disclosure.
• Informed consent is required under Rule 1.6 because client information will be input into the tool.
• GAI use is relevant to the fee basis under Rule 1.5.
• The tool's output will influence a significant decision in the representation.

"[L]awyers must disclose their GAI practices if asked by a client how they conducted their work, or whether GAI technologies were employed in doing so, or if the client expressly requires disclosure under the terms of the engagement agreement or the client's outside counsel guidelines."
Opinion 512, at 8.

Even where Rule 1.4 does not require disclosure, the Opinion notes that a firm may describe its GAI practices in the engagement letter as a form of effective communication. Most firms do this with a short, plain-language paragraph that names the tool categories used and invites the client to request additional detail.

Rule 1.5: Fees and billing

Rule 1.5 requires reasonable fees and expenses. Opinion 512 applies this rule to GAI in two ways that carry direct operational consequence.

First, efficiency gains from GAI cannot be billed at the pre-GAI hourly rate. The Opinion states the rule:

"GAI tools may provide lawyers with a faster and more efficient way to render legal services to their clients, but lawyers who bill clients an hourly rate for time spent on a matter must bill for their actual time."
Opinion 512, at 12.

The same principle extends to flat and contingent fees. If a flat fee was set against pre-GAI assumptions about time required, the Opinion signals that the fee may no longer be reasonable if GAI materially compresses the work.

Second, learning time cannot be billed to the client:

"[A] lawyer may not charge a client to learn about how to use a GAI tool or service that the lawyer will regularly use for clients because lawyers must maintain competence in the tools they use, including but not limited to GAI technology."
Opinion 512, at 14.

The Opinion also distinguishes overhead-type tool costs from client-specific tool expenses. An example of overhead: a grammar-check feature bundled into word processing. An example of a pass-through expense: a per-matter charge from a contract-review vendor. Firms should map each approved tool to one of those two categories and disclose the billing treatment to clients in advance.

Rules 5.1 and 5.3: Supervisory duties

Under Rules 5.1 and 5.3, managerial and supervisory lawyers are responsible for the conduct of other lawyers and nonlawyer assistants. Opinion 512 places GAI tools within this framework. Internally, managing partners must establish firm-wide policies. Supervisory lawyers must train and oversee staff use. The Opinion's formulation:

"Managerial lawyers must establish clear policies regarding the law firm's permissible use of GAI, and supervisory lawyers must make reasonable efforts to ensure that the firm's lawyers and nonlawyers comply with their professional obligations when using GAI tools."
Opinion 512, at 10.

Externally, the Opinion applies Rule 5.3(b)'s reasonable-efforts duty to the GAI tool's provider. It carries forward the diligence framework from the ABA's prior cloud-computing and outsourcing opinions. For GAI vendors, that framework covers: reference checks and credentials; security policies and protocols; confidentiality agreements; conflicts screening where applicable; and whether the provider retains or claims proprietary rights to submitted content.

The documentation that follows from Rules 5.1 and 5.3 is concrete. Keep a written firm AI policy (see the policy template). Maintain training completion records for each attorney and staff member, a vendor due diligence file for each approved tool, and a supervision protocol for AI-assisted work product.

Rule 3.3: Candor toward the tribunal

Rule 3.3 prohibits false statements of law or fact to a tribunal. It also requires remedial action if a lawyer learns that material evidence was false. Dishonesty, fraud, deceit, and misrepresentation fall under Rule 8.4(c). Opinion 512 applies both rules to GAI-assisted filings. On the risk, the Opinion is direct:

"Even an unintentional misstatement to a court can involve a misrepresentation under Rule 8.4(c). Therefore, output from a GAI tool must be carefully reviewed to ensure that the assertions made to the court are not false."
Opinion 512, at 10.

The scope of the pre-filing review duty is broader than just citation checking:

"In judicial proceedings, duties to the tribunal likewise require lawyers, before submitting materials to a court, to review these outputs, including analysis and citations to authority, and to correct errors, including misstatements of law and fact, a failure to include controlling legal authority, and misleading arguments."
Opinion 512, at 10.

This is where the rule meets the sanctions cases. Courts have applied Rule 3.3 (or its state analogs) to AI-generated hallucinations and imposed sanctions ranging from modest fines to fee awards exceeding $1.5 million. Referrals to state bar discipline have followed in the more serious cases.

Notable sanctions cases in this line:

• Mata v. Avianca, Inc. (S.D.N.Y.) : $5,000
• Wadsworth v. Walmart Inc. (D. Wyo.) : $3,000 (Ayala) + $1,000 each (Morgan, Goody); Ayala's pro hac vice revoked
• Garner v. Kadince (Utah Ct. App.) : Attorney fees + client fee refund + $1,000 to access-to-justice nonprofit
• Mezu v. Mezu (Md. App. Ct.)
• Johnson v. Dunn (N.D. Ala.)
• Jason M. Hatfield, P.A. v. Pirani (W.D. Ark.) : $1,578,172 attorney fees + $93,388 costs
• People v. Crabill (Colo. OPDJ) : 1 year + 1 day suspension; 90 days active
• Shahid v. Esaam (Ga. Ct. App.) : $2,500
• Kohls v. Ellison (D. Minn.)
• Creech v. City of Raleigh (N.C. Ct. App.)

Opinion 512 compliance checklist

Twelve items, each mapped to the rule it documents. If a firm can answer yes to every item, it holds the documentation a malpractice carrier or disciplinary investigator is likely to request.

1. Written AI policy in force (Rules 5.1, 5.3). The firm has a current policy, dated within the last 12 months, naming approved tools, prohibited uses, and supervisory responsibility.
2. Approved-tools list, reviewed (Rules 1.1, 1.6). Each tool on the list has a completed vendor review covering Terms of Use, privacy policy, data retention, and training-on-input behavior.
3. Prohibited-tools list (Rule 1.6). The policy identifies consumer tools (including personal ChatGPT, Gemini, Claude, and similar) that are not approved for firm work, and explains why.
4. Tool-specific informed consent language (Rule 1.6). For any self-learning tool into which client information will be input, the firm has client-facing consent language that meets the Opinion's "not boiler-plate" standard.
5. Engagement letter disclosure (Rule 1.4). The firm's default engagement letter describes its AI use in plain language, or documents the decision not to disclose and the reasoning.
6. Training records (Rules 1.1, 5.1, 5.3). Every attorney and staff member using an approved tool has completed tool-specific training, logged by date.
7. CLE tracking for technology competence (Rule 1.1). The firm tracks ongoing CLE or equivalent training so that technological competence is maintained as tools change.
8. Pre-filing verification protocol (Rule 3.3). Any filing that used AI research assistance is logged, with the citation-verification step documented by the attorney who reviewed it.
9. Billing policy for AI time (Rule 1.5). The firm has written guidance on billing AI-assisted work: hourly rules, flat-fee treatment, overhead vs. pass-through expense classification for each approved tool.
10. Supervision protocol for AI-assisted work product (Rule 5.1, 5.3). Any work product generated with AI assistance is reviewed by a supervising lawyer before it leaves the firm, with the review logged.
11. Incident response procedure (Rule 1.6). The firm has a written procedure for responding to an AI-related confidentiality incident, including client notification and bar reporting where applicable.
12. Annual policy review (Rules 1.1, 5.1). The firm reviews and re-dates the AI policy at least annually, and on any material tool change.

State cross-reference

Opinion 512 is the national baseline. Several states have issued their own formal opinions or practical guidance that add to, clarify, or diverge from the ABA's framing. Firms should read Opinion 512 alongside their state's guidance. More detailed state-level sources to date:

• Florida Bar Ethics Opinion 24-1 (January 2024), which predates Opinion 512 and is repeatedly cited in it.
• North Carolina State Bar 2024 FEO 1 (adopted 2024), the first formal NC opinion on GAI.
• California State Bar Practical Guidance on Generative AI (November 2023), cited in Opinion 512 as a source on confidentiality.
• Pennsylvania and Philadelphia Joint Formal Opinion 2024-200, which Opinion 512 cites on cross-representation leakage in self-learning tools.
• New York, Texas, Illinois, and other state bars that have issued AI-specific guidance since 2024.

The state tracker carries primary-source citations for each state's guidance.

Next steps for your firm

• If your firm does not yet have a written AI policy, start with the law firm AI policy template, which is mapped to Opinion 512 section-by-section.
• If your malpractice renewal is approaching, our AI Liability Insurance for Law Firms guide explains the current carrier landscape: silent AI cover, named exclusion endorsements, and the affirmative-cover market. It also gives the seven questions to put to your underwriter at renewal. For deeper carrier-side reference, see our carrier documentation guide, which includes FY2025 10-K language and form-by-form coverage mechanics.
• Check your state's guidance on the state tracker. Where state rules are stricter than Opinion 512, the state rule controls.
• Opinion 512 is rule-by-rule ethics guidance; it does not cover the operational AI risk-management infrastructure that bar opinions and carriers increasingly expect. For that operational layer, see the NIST AI Risk Management Framework and its Generative AI Profile. Opinion 512 does not cite either NIST publication by name; the connection is operational rather than textual.

Primary source

ABA Standing Committee on Ethics and Professional Responsibility, Formal Opinion 512, Generative Artificial Intelligence Tools (July 29, 2024). Full text: americanbar.org (PDF) .

Last verified against primary source: 2026-04-24.