Guide

AI receptionists for law firms: what they handle, the rules, and when you need a person

What AI phone and chat receptionists do well, what the ABA and state bars say about them, the disclosure laws to check, and when a person should answer.

Short answer: an AI receptionist can answer every call instantly, at any hour, ask the same questions every time and book appointments. It can’t use judgment, and the rules around it are still forming. If you use one, tell callers they’re talking to AI, keep it away from legal advice, protect what prospective clients tell it, and make it easy to reach a person. Many firms will get the best result from both: a person during business hours and for Spanish-speaking callers, and AI for overflow and after hours.

Who wrote this, and what it isn’t. Legal Staffing Pros places human intake staff with law firms, so we have a stake in this question; we’ve tried to be fair to the software. This is general information, not legal advice: the rules below differ by state and change often, so check them with your ethics counsel before you switch anything on.

What AI receptionists do well

  • Instant answers at any hour. No hold music, no voicemail, no busy signal when three people call at once.
  • The same questions, every time. A well-configured system follows the script without skipping steps.
  • Booking and routing. Many connect to calendars and case management systems, and can send a call to the right person.
  • A written record. Transcripts and summaries make every inquiry reviewable.

For simple, repeatable calls, and for the hours when no one is at the desk, those are real advantages.

Where a person still matters

  • Judgment. Deciding that a story doesn’t add up, that a caller is describing two different matters, or that an attorney should call back today is hard to script.
  • Distressed callers. People calling a law firm are often injured, frightened or angry. Whether they feel heard can decide whether they book.
  • Spanish and code-switching. Many callers move between Spanish and English mid-sentence, and the details that matter, like dates and what a supervisor said, are the easiest to lose.
  • Conflicts and represented parties. Noticing that the caller already has a lawyer, or that the other side is a current client, needs to happen before the story is told.

Callers notice the difference. In a Gartner survey of 3,566 customers in February and March 2026, 87% said it is essential for companies using generative AI in customer service to offer a way to reach a human agent, while 50% said their interactions are easier when companies use it.

What the ethics rules say

ABA Formal Opinion 512 (July 29, 2024) is the ABA’s guidance on lawyers’ use of generative AI. The parts that matter most for intake:

  • Confidentiality. Before client information goes into a tool that may learn from or share what it receives, the client’s informed consent is required, and boilerplate in an engagement letter isn’t enough. Rule 1.18 gives prospective clients similar protection, so what callers tell an AI receptionist deserves the same care.
  • Supervision. Lawyers with managerial authority should set clear policies on how AI may be used, and supervising lawyers must make reasonable efforts to see that they’re followed, under Rules 5.1 and 5.3.

Opinion 512 is about lawyers’ use of AI tools in general. On intake chatbots specifically, Florida Bar Ethics Opinion 24-1 (January 19, 2024) is more direct. It says a lawyer using an AI chatbot for intake must inform prospective clients that they are communicating with an AI program, that the chatbot should not give legal advice and should refer legal questions to the lawyer, and that the lawyer should consider screening questions that limit its communications, for example with people who already have a lawyer. The lawyer stays responsible if the chatbot gives misleading information or communicates in a way that is intrusive or coercive.

Disclosure and calling laws to check

Beyond the ethics rules, a growing number of state laws cover AI that talks to consumers. Some examples, current when we checked them; your state may have others:

LawWhat it requiresWhy it matters for intake
Utah, Artificial Intelligence Consumer Protection Amendments (effective May 7, 2025)A business using generative AI with consumers must disclose it when a person clearly asks whether they’re talking to a human. Separate up-front disclosure applies to occupations licensed through Utah’s Department of Commerce in “high-risk” interactions, which include legal advice.Configure the system to answer truthfully when asked, and ask counsel whether the up-front rule reaches your firm
Maine, Act to Ensure Transparency in Consumer Transactions Involving AI (signed June 12, 2025)An AI chatbot used in a way that could mislead a reasonable consumer into thinking they’re dealing with a human must clearly and conspicuously disclose that it isn’t.For voice, that means saying so at the start of the call
California, Business and Professions Code section 17941It is unlawful to use a bot online to mislead a person about its artificial identity in order to encourage a sale, unless the bot clearly discloses that it is a bot.Relevant to website chat that encourages people to hire the firm
FCC declaratory ruling (February 2024)AI-generated voices count as an “artificial” voice under the Telephone Consumer Protection Act, so calls using them need the called party’s prior express consent unless an exemption applies.An AI that calls leads back is placing calls, not just answering them
California Penal Code section 632Recording a confidential communication without the consent of all parties is a crime.Systems that record or transcribe calls need a consent notice at the start

A setup that uses both

If you add AI to intake, these choices keep it on the right side of the rules above and keep callers from slipping away:

  1. Disclose first. The greeting says the caller is speaking with an automated assistant, and that the call is recorded if it is.
  2. One step to a person. Saying “a person” or pressing zero reaches someone during business hours, or books a callback at a stated time after hours.
  3. No legal advice. The system collects facts and books appointments. Questions about the person’s case go to an attorney.
  4. Names before the story. Collect names for the conflict check first, and keep the initial questions to what the firm needs to decide whether to take the matter.
  5. Spanish to a Spanish speaker. Route callers who prefer Spanish to a bilingual person whenever one is on duty.
  6. Firm-controlled data. Know where recordings and transcripts are stored, who can see them, and whether the vendor uses them to train its models.
  7. Weekly review. A named person reads a sample of transcripts every week and fixes what went wrong.

Checklist before you switch AI intake on

AI intake checklist

  1. Ask your ethics counsel which rules apply in each state where you take calls
  2. Get the vendor’s terms in writing: where data is stored, who can access it, and whether it trains on your calls
  3. Write the greeting so it says the caller is talking to an automated assistant
  4. Add a recording notice if calls are recorded or transcribed
  5. Set the system to answer truthfully when asked whether it is a person
  6. Block legal advice; route case questions to an attorney
  7. Collect conflict-check names before the caller’s story
  8. Give callers one step to a person, or a callback at a stated time
  9. Route Spanish-speaking callers to a bilingual person when one is on duty
  10. Don’t use it to place outbound calls without checking consent rules
  11. Name the person who reviews transcripts every week
  12. Write the policy down and train staff on it

Where a dedicated person fits

A dedicated intake specialist covers what AI handles least well: judgment, distressed callers and full conversations in Spanish. At Legal Staffing Pros, a full-time bilingual intake specialist is $2,495 a month, and evening or Saturday coverage adds $200. Many firms pair one with an AI system or an answering service for nights. For the full intake process, see our complete guide to legal intake.

Sources

ABA Formal Opinion 512, Generative Artificial Intelligence Tools, July 29, 2024, as summarized by the ABA Section of Real Property, Trust and Estate Law. Florida Bar Ethics Opinion 24-1, January 19, 2024, as reported by the ABA Journal and Hinshaw & Culbertson. Gartner, press release, August 4, 2026. Utah S.B. 226 (2025), enrolled. Maine: Verrill summary of the Act. California Business and Professions Code section 17941 and Penal Code section 632. FCC, Declaratory Ruling FCC 24-17, adopted February 2 and released February 8, 2024. All read November 15, 2026.

Tell us who you need. We’ll show you who’s available.

A short call covers the role, the hours, your software and your budget. You interview everyone before anyone starts, and you pay nothing until they do.