An AI receptionist for law firms: intake without giving advice

A law firm's AI receptionist runs intake. It captures the matter type, the names needed for a conflict check, the urgency and the contact details, and books the consultation. It gives no legal advice, does not predict outcomes, and says clearly that nothing in the conversation creates an attorney-client relationship. Those limits are the design, not a restriction on it.

What intake it can do

Most first contacts with a small firm are the same short set of facts: what kind of matter, roughly what happened, whether there is a deadline, who the other party is, and how to reach the person. The AI asks those in your order, in plain language, and stops when it has them. It then books the consultation with the right attorney based on practice area, or routes the summary to whoever screens new matters.

The conflict-check questions

Before a consultation is booked, the AI asks for the names of the other parties involved. It does not run the check; it captures the names cleanly so your team can, and the appointment is marked as pending until they do. This is a small thing that saves an awkward cancellation later.

Urgency goes to a person now

A court date this week, an arrest, a served document with a deadline, an eviction notice, a protective order: these are escalated to a person immediately with the details captured so far. The AI does not attempt to advise on what to do before the attorney calls, and it does not say "you should be fine".

What it must not say

No opinion on the merits. No estimate of what a case is worth. No "yes, you have a claim" or "that is illegal". No explanation of the law as it applies to their facts. When asked, it says that an attorney will discuss that at the consultation, and it means it.

Confidentiality and what to collect by text

Text is not the place for the story. The AI asks for enough to route the matter and books the consultation; the detailed account happens with the attorney. It does not ask for Social Security numbers, account numbers or documents by text, and if a prospective client sends them anyway, the conversation is flagged for a person.

Disclaimers, in your words

Load the exact disclaimer language your firm uses: that the conversation is with an automated assistant, that it does not create an attorney-client relationship, and that no advice is being given. The AI includes it at the start and repeats it when relevant. Your state bar's rules on advertising and intake apply to what the AI says; review the wording with that in mind.

Frequently asked questions

Is AI intake allowed under bar rules?

Rules vary by state and are changing. Many firms use automated intake today; what matters is that no advice is given, disclaimers are clear, and confidentiality is protected. Check with your state bar. This post is not legal advice.

Can it take a retainer payment?

Not by text. It sends the link to your payment page and tells the client what to expect. Card numbers are never collected in a conversation.

Can it handle Spanish-speaking callers?

Depends on the product. Ask whether the AI answers in the language the client writes in and whether your approved knowledge is translated or answered from English.

What Frontiva does here

The legal page walks through intake for a small firm. In Frontiva every new enquiry becomes a lead in a pipeline with a visible score, an owner and follow-up tasks with due dates (lead management), and the escalation keywords you set (court, arrest, deadline, served) hand a conversation to a person before any answer is attempted.

For the wider picture, and the line between answering and advising, see what is an AI receptionist.