Fred: Client Intake That Never Crosses Into Legal Advice

A visitor lands on your firm’s site late at night and types out their situation: a contract dispute, a deadline they think is coming, a question about whether they have a case. A generic chatbot, built to be helpful, answers. In a few sentences it has given legal advice, implied an attorney-client relationship, and possibly gotten the deadline wrong, under your firm’s name and your bar number. The firm wanted a tool to book consultations. It got one that practices law without a license.

The threat and standard articles in this series lay out how that exposure builds. This piece is the answer: what a compliant AI assistant for law firms actually looks like, and why the design, not the disclaimer at the bottom of the page, is what keeps the firm clear of an ethics problem.

The Real Choice Is Governed or Ungoverned

Firms tend to ask whether they should use AI on the site at all. Prospective clients have already decided for them, they expect an instant response, and the firm that makes them wait loses the intake to one that did not. The question that matters is whether the assistant is governed.

An ungoverned chatbot answers the "do I have a case" question, speculates on a statute of limitations, and reassures a stranger in a way that reads as legal advice. A governed one captures the same intake and routes every one of those questions to an attorney. Identical responsiveness, entirely different exposure under the rules of professional conduct.

How Fred Holds the Line

Fred starts from the opposite default of a general-purpose chatbot. A generic bot answers anything unless instructed otherwise; Fred answers only what it is cleared to, and the boundary is enforced by the system rather than written into a prompt it can wander away from.

In practice, Fred works from your firm’s own content, captures the intake, and routes anything that constitutes legal advice, case evaluation, or a deadline question to an attorney. It does not opine on a matter, because doing so risks the unauthorized practice of law and the supervisory duty Rule 5.3 places on the nonlawyer help a firm uses. It is careful with what visitors share, mindful of the confidentiality duty in Rule 1.6, and it does not make the kind of claims about results or expertise that Rule 7.1 prohibits. The ABA’s first formal ethics guidance on generative AI made clear these duties follow the tool; Fred is built around that fact, and it informs Fred’s legal guardrail pack, one of more than 50 industry packs.

The design proves itself under pressure. A worried visitor will rephrase "do I have a case" several ways, and a prompt-instructed bot eventually answers the version it was not warned about. Fred does not rely on catching the phrasing. It decides what may be said before the reply exists, so the advice answer never reaches the visitor.

Fred vs. a Generic AI Chatbot

The table is the whole argument on one screen. A generic tool is helpful until helpful becomes the practice of law. Fred is helpful across everything that carries no ethics risk and structurally silent everywhere that does.

What Your Firm Actually Gets

Look past the rules at the business case. Fred answers the routine questions that flood a firm’s inbox, practice areas, how a consultation works, what to bring, hours and location, and it does it instantly, after hours, in the visitor’s own words. It captures the intake with context, so the attorney who follows up already knows the type of matter and the urgency. Everything that requires judgment, or that could be construed as advice, stays with a lawyer.

The intake quality is the quiet advantage. Instead of a bare contact form, the attorney opens a qualified summary, what the visitor came in for, which practice area, how time-sensitive it feels, captured the moment the person was motivated enough to ask. That context is the difference between a callback that converts to a retained client and one that goes cold.

So the question is not whether your competitors will put AI on their sites. They will. It is whether yours is governed before a prospective client asks it the one question it should never answer.

Frequently asked questions