ABA Formal Opinion 512, issued on July 29, 2024, is the American Bar Association's first formal ethics opinion on generative AI. It now comes up in almost every conversation about AI and law-firm marketing. Read the opinion itself, though, and you find it is not an advertising opinion. This piece covers what it does say, the one footnote where marketing comes in, and what that means when the agency writing your website uses AI.
What Opinion 512 covers
The 15-page opinion is about lawyers using generative AI tools in their practice. Its opening summary lists the duties involved: competence, protecting client information, communicating with clients, supervising employees and agents, advancing only meritorious claims, candor toward the tribunal, and charging reasonable fees.
| Duty | Model Rule | What Opinion 512 says, in short |
|---|---|---|
| Competence | 1.1 | Lawyers need a reasonable understanding of a tool's capabilities and limits, and should review AI output for accuracy. |
| Confidentiality | 1.6 | Before putting client information into a self-learning AI tool, the lawyer needs the client's informed consent. |
| Communication | 1.4 | In some situations, lawyers should tell clients they are using AI tools. |
| Meritorious claims and candor | 3.1, 3.3 | Court filings must be checked; fabricated citations and misleading arguments remain the lawyer's responsibility. |
| Supervision | 5.1, 5.3 | Managing lawyers should set clear firm policies on AI use and make reasonable efforts to ensure staff and outside providers comply. |
| Fees | 1.5 | Fees must stay reasonable. In most cases a lawyer can't bill a client for time spent learning an AI tool. |
Where marketing comes in: one footnote
Opinion 512 does not analyze lawyer advertising. In a footnote, it says additional issues may arise under other rules, and it names Model Rule 7.1, the rule against false or misleading communications about a lawyer's services, along with the conflict rules. That footnote points to state opinions that go further, including a Florida Bar ethics opinion (24-1) discussing AI chatbots under Florida's advertising rules.
So the honest summary is this. Opinion 512 does not create new marketing rules. Rule 7.1 and your state's version still decide whether a website page, a directory profile or an AI-written FAQ is misleading. What Opinion 512 adds is the reminder that a lawyer can't hand that judgment to a tool, and that the supervision duties reach the outside providers a firm relies on.
What it means when an agency writes with AI
Two parts of the opinion matter most for marketing work.
- Supervision of outside providers. The opinion draws on earlier guidance about outsourcing: check the vendor's credentials, understand its security practices, and use confidentiality agreements. It says those concepts apply to generative AI providers too. A marketing agency is an outside provider. If it drafts your content with AI, your firm still needs to know how that content is checked.
- Client information. Marketing rarely needs confidential client information, and it shouldn't get it. Case studies, testimonials and results claims are where that line is easiest to cross. Under the confidentiality analysis, putting client details into a self-learning tool needs informed consent, so the simplest policy is to keep client matter information out of marketing AI tools entirely.
Everything else about the content itself still runs through Rule 7.1: no unsubstantiated "best" claims, no implied guarantees, no "specialist" wording your state doesn't allow, and results stated with the context your bar requires. Our guide to Rule 7.1 and AI search covers those failure modes, and the state-by-state comparison covers how Florida, Texas, New York, New Jersey and Louisiana add to it.
Five questions to ask any agency that uses AI
- Who reviews AI-drafted content before it publishes, and against which rules: Rule 7.1 only, or your state's rules too?
- Does any client or matter information go into the AI tools you use? If so, which tools, and on what terms?
- How do you check facts the AI produces, such as attorney credentials, bar admissions and results?
- Who at our firm approves content before it goes live, and how is that approval recorded?
- What happens when a published page turns out to be wrong?
A good answer to all five is a written workflow, not a promise. None of this guarantees how any AI engine will describe your firm; it keeps what you publish defensible.
The short version
Opinion 512 is about how lawyers use AI, not about what lawyers may advertise. For marketing, it adds two things: supervise the outside providers who use AI on your behalf, and keep client information out of tools that could disclose it. Whether a given page is misleading is still a Rule 7.1 question, answered under your state's rules.
Sources: ABA Formal Opinion 512 (July 29, 2024) and the ABA's announcement, both read October 16, 2026. This is general information for marketers, not legal advice; confirm requirements with your state bar or your supervising attorney.
About Legal Torch. Legal Torch is a generative engine optimization (GEO) agency that works only with law firms. We screen every public-facing deliverable against ABA Model Rule 7.1 and the firm’s state advertising rules before it publishes, and the firm’s attorney approves it. We measure AI visibility on each engine separately and never guarantee outcomes. See our GEO service and how the compliance review works.
