AI-Generated Attorney Bios and Practice Area Pages: Does NJ RPC 7.1 Have a Problem With What You're Publishing?

Most NJ solo attorneys who've used ChatGPT or Claude to draft a website bio or practice area page did it on a Sunday afternoon when they finally got tired of staring at a blank screen. The output looked polished. The grammar was clean. They made a few edits, copied it into their CMS, and moved on.

That workflow isn't inherently wrong. But the specific risks hiding inside AI-generated attorney marketing copy are different from the risks in, say, AI-generated discovery responses, and most small firms haven't thought through them carefully.

Here's what actually goes sideways.

The Credentialing Problem

AI models are trained to write convincingly, which means they default to the kinds of credential signals that look good in attorney bios. If your prompt says "write a bio for a NJ family law attorney with 12 years of experience," the model may produce language like "a proven track record of successful outcomes in high-stakes custody disputes" or "recognized for her strategic approach to complex asset division."

You didn't write that. You also didn't fact-check it.

Under RPC 7.1, a communication about a lawyer's services is prohibited if it contains a material misrepresentation of fact or law, or omits a fact necessary to make the statement considered as a whole not materially misleading. "Proven track record" and "recognized for" are vague enough that they might slide past a first read, but if the attorney hasn't actually been recognized for anything and has a modest litigation record, those phrases are doing real work in creating a false impression. That's the definition of misleading under the rule.

The NJ Office of Attorney Ethics has historically taken RPC 7.1 seriously. The rule covers websites, social media profiles, and any other communication about services. It's not limited to formal ads.

Superlatives and Comparison Claims

AI marketing copy loves superlatives. Without specific instructions, models often produce phrases like "one of the leading estate planning attorneys in New Jersey" or "among the top-rated small firm litigators in Bergen County." These comparison claims are specifically addressed in NJ's advertising guidance. The Committee on Attorney Advertising has flagged that comparative statements require a factual basis, a ranking, a verifiable methodology, a published rating. A ChatGPT output doesn't supply one.

The fix is simpler than it sounds: add a line to your AI prompt that says something like "avoid all comparative claims, superlatives, or outcome guarantees, and do not imply any recognition or awards unless I specify them." Most attorneys don't think to do this, because they're prompting for speed, not compliance.

Practice Area Pages and the UPL Adjacent Problem

This one is subtler. AI-generated practice area pages often include explanatory legal content, a paragraph on how NJ divorce law handles equitable distribution, or how a DUI first offense is typically resolved in municipal court. That content is usually accurate at a general level, but it can cross into territory that functions like legal advice if a prospective client reads it, relies on it, and acts on it without retaining anyone.

This isn't a direct RPC 7.1 violation, but it creates a confluence of risks: potential UPL-adjacent issues if the firm is providing substantive guidance through marketing copy, and a duty-of-care question if a reader is harmed by relying on general AI-generated content that doesn't apply to their specific facts. A short disclaimer at the bottom of every practice area page, "this content is for general informational purposes only and does not create an attorney-client relationship", is basic hygiene, but AI-generated pages frequently omit them.

A Practical Review Protocol

Before publishing any AI-drafted marketing copy, run it through four specific questions:

1. Credential check. Does this copy imply any award, ranking, recognition, or outcome pattern that I can't verify with a source?
2. Comparative language check. Does this copy position me as superior to, or more capable than, other attorneys in any way that lacks a factual basis?
3. Outcome language check. Does this copy suggest or imply specific results, even subtly, in violation of NJ's prohibition on outcome-based advertising?
4. Disclaimer check. If this page includes substantive legal content, does it carry an appropriate informational disclaimer?

The NJ Committee on Attorney Advertising will review your website if a grievance is filed. What reads as "just marketing language" in draft can look quite different in a formal inquiry.

The smarter use of AI for attorney marketing isn't writing the final copy, it's drafting a first pass that you then rewrite with your own voice and review against these four questions. That step takes 20 minutes. An OAE inquiry takes considerably longer.