When an NJ Attorney's AI Marketing Copy Crosses the Line: What RPC 7.1 Actually Prohibits

Solo attorneys and small-firm practitioners in New Jersey have started using AI to write their own marketing content. Website bios, practice area pages, LinkedIn posts, client intake emails, all of it can now be drafted in minutes. The quality is often surprisingly good, and the time savings are real.

The problem is that AI writes to impress. That's its default mode. And attorney advertising in New Jersey operates under rules that make "impressive" language a liability if it isn't also accurate.

RPC 7.1 prohibits a New Jersey attorney from making a false or misleading communication about the attorney or the attorney's services. That sounds simple. In practice, with AI-generated copy in the mix, it creates more friction than most attorneys anticipate.

How AI Copy Tends to Violate RPC 7.1

When you prompt an AI tool to write your law firm's practice area page, it will almost certainly produce language that includes superlatives, implied results, and vague claims of specialized expertise. That's not a flaw in the prompt, that's how the tools are trained. They're optimizing for engagement and conversion, not Bar compliance.

Here's what that looks like in practice:

A family law attorney in a suburban NJ county asks an AI to write her divorce practice page. The output describes her as "a trusted advocate who has helped hundreds of families achieve favorable outcomes." That single sentence has at least two potential RPC 7.1 problems. First, "favorable outcomes" implies a track record of results without the disclaimer NJ advertising rules require for result-based claims. Second, "trusted advocate" is the kind of subjective characterization that the rule expressly identifies as potentially misleading if not accompanied by something substantive to back it up.

The NJ Rules of Professional Conduct, read alongside the state's attorney advertising regulations under the Court Rules, require that any communication about a lawyer's services be truthful and not create unjustified expectations about results. The Office of Attorney Ethics (OAE) has historically taken a close look at unsubstantiated superlatives in attorney advertising, even when the attorney's intent was clearly benign.

The Specific Phrases That Should Trigger a Review

If your AI-drafted marketing content includes any of the following, it warrants a hard look before publishing:

• Any variation of "experienced" or "seasoned" paired with a specific practice area, unless you can substantiate the claim with years in practice or volume of matters handled.
• Language suggesting a specialty or certification in a field (e.g., "specialist in NJ criminal defense") when you don't hold a designation recognized under New Jersey's certification program through the Supreme Court.
• Testimonial-style language, even if the AI invented it as a placeholder, it can slip through into a draft that gets published without review.
• References to past case results without the required disclaimer that prior results do not guarantee future outcomes.

None of these are obscure traps. They're the same issues NJ attorneys have always faced with marketing vendors who write ad copy without understanding Bar rules. AI is just a faster version of that vendor.

What a Practical Review Workflow Looks Like

The solution isn't to stop using AI for marketing content. The efficiency gain is worth preserving. The answer is building a short compliance step into the publishing workflow.

Before any AI-generated marketing copy goes live, on a website, in a directory listing, in a social post, or in an email campaign, it should run through a checklist tied directly to RPC 7.1 and New Jersey Court Rule 1:39A (which governs attorney specialty certifications). That checklist doesn't need to be long. It needs to ask five targeted questions:

1. Does this copy claim or imply a specialty or certification the attorney doesn't hold under the NJ Supreme Court's certification program?
2. Does it reference past results without a compliant disclaimer?
3. Does it use a superlative (best, top, most experienced, leading) that cannot be substantiated?
4. Does it include any language that could create a reasonable expectation of a specific outcome?
5. Was any of it generated using a client's actual matter details, even anonymized ones, without consent?

That last question matters more than it might seem. Attorneys who prompt AI tools with real client facts to generate "case study" style marketing content are creating a confidentiality issue on top of an advertising issue.

A Word on Social Media Specifically

LinkedIn and Instagram are where I see NJ solo attorneys getting into the most trouble with AI-generated content. The short-form format encourages bold claims, and AI optimizes for engagement signals. A post that says "Another win for a client facing impossible odds" is doing real advertising work under RPC 7.1, even if the attorney thought of it as just a social media update.

The New Jersey OAE has issued guidance treating social media posts as attorney advertising when they are directed at prospective clients or describe the attorney's services. That's a broad category. When in doubt, apply the same RPC 7.1 review to social content that you'd apply to a formal website page.

Using AI to draft attorney marketing content is a sound operational decision. Publishing that content without a compliance review is the part that creates exposure. Adding fifteen minutes of structured review before anything goes live is a small investment against a problem that, once it reaches the OAE, is never small.