You Hired a Legal AI Tool. Did You Just Accidentally Advertise Your Practice?

There's a category of AI risk that almost no New Jersey solo or small firm attorney is thinking about — and it's not hallucinated citations or confidentiality breaches.

It's advertising compliance.

Specifically: the moment your AI tool drafts, publishes, or sends any client-facing communication on your firm's behalf, it may be engaging in attorney advertising under NJ RPC 7.1 and RPC 7.3. And because it happened automatically, without attorney review, there's a real chance you have no idea it already occurred.

---

How This Happens in Practice

Consider a few scenarios that are increasingly common in solo and small firm tech stacks:

• Your intake chatbot (powered by an AI tool like Lawmatics, Clio Grow, or a custom GPT) sends a prospective client an automated follow-up message summarizing your "areas of expertise" and inviting them to schedule a consultation.
• Your AI email assistant drafts and queues a response to a cold inquiry referencing your track record and "successful outcomes" in family law cases.
• Your AI-powered website tool automatically updates your practice page bio with new language generated from your recent case history data.

Each of these interactions is a communication "concerning a lawyer's services" under RPC 7.1. And if any of them include a direct solicitation to a prospective client — particularly one who may be in a vulnerable legal situation — you've wandered into RPC 7.3 territory, which in New Jersey carries specific restrictions around real-time contact and targeted solicitation.

The AI didn't know that. It was just doing its job.

---

What RPC 7.1 Actually Requires of AI-Generated Content

RPC 7.1 prohibits false, misleading, or deceptive communications about a lawyer or the lawyer's services. It's straightforward on its face — but AI tools introduce a subtle compliance trap: generated language that sounds confident and specific is often neither verifiable nor accurate.

When an AI drafts a bio line like "Trusted by hundreds of New Jersey families in high-stakes custody disputes," it may be pulling from a prompt you gave it months ago, extrapolating from client volume estimates, or simply pattern-matching against what sounds compelling. Unless you personally reviewed and verified that claim, it's potentially false or misleading under the Rule — and you are responsible for it, regardless of who (or what) wrote it.

Under RPC 5.3, your obligation to supervise non-lawyer assistance extends to automated tools. The fact that the output came from software rather than a paralegal is not a defense.

---

The RPC 7.3 Angle Is Even More Specific

RPC 7.3 governs direct contact with prospective clients. New Jersey's version — aligned with post-2023 amendments to the ABA Model Rules — permits written solicitation in many contexts but prohibits contact when the prospective client has made known a desire not to be solicited, or where the contact involves coercion, duress, or harassment.

The practical problem: AI-powered intake workflows often don't check any of this. They fire follow-up sequences based on behavioral triggers (opened an email, visited a pricing page, abandoned a form), not on legal eligibility to receive solicitation. If someone visited your site after a serious personal injury and receives three automated AI-drafted follow-up messages before they've had any human contact with your office, that sequence deserves scrutiny.

---

Your Three-Part Compliance Audit

If you're running any AI-assisted client communication in your practice — even lightly — here's where to start:

1. Map every automated touchpoint. List every place where AI generates content that reaches a prospective or current client: chatbots, intake sequences, email automations, website content generators, scheduling tools with auto-messaging. If it sends or publishes something on your behalf, it's on this list.

2. Apply the RPC 7.1 truth test to stored templates and generated content. Pull the last 10 client-facing messages your AI tools generated. Read each one as if you were a skeptical bar investigator. Are there outcome claims? Superlatives? Characterizations of your experience that you can't immediately verify? Flag and revise.

3. Build a human review gate for solicitation messages. Any AI-generated message that constitutes a first outreach to a prospective client — particularly in practice areas involving personal injury, criminal defense, family law, or immigration — should require attorney review before it sends. This is not optional under RPC 5.3. It is also, frankly, just good judgment.

---

The Deeper Issue for Solo Practitioners

Large firms have marketing compliance teams. You don't. That's precisely why automation is appealing — and precisely why the risk lands squarely on you.

The New Jersey Office of Attorney Ethics has not issued specific guidance on AI-generated advertising content as of this writing. That silence is not protection. RPC 7.1 and 7.3 apply to the communication, not to the method used to create it.

The good news: this is one of the more fixable AI compliance problems in a solo practice. It requires awareness, a brief audit, and one new operational habit — reviewing what your tools are saying on your behalf before they say it.

Your AI is a capable communicator. Make sure it's not communicating you into a bar complaint.