AI legal intake means an AI agent answers new-client calls, chats and web forms for your firm, collects the facts your intake team needs, screens for practice area and jurisdiction, gathers the names you need for a conflict check and books a consultation with the right attorney. It doesn’t give legal advice, assess the strength of a case or promise representation. Those stay with lawyers, and the ethics questions around using AI for intake are answered by your state’s rules, not by a software vendor.
This guide covers how intake automation works in a law firm, where the line sits between intake and advice, and which professional conduct topics to review with your state bar’s rules before you launch.
Why intake is where firms lose clients
For many practices, especially personal injury, family law, criminal defense, immigration and employment, a new matter starts with a phone call made at a stressful moment. The person calling may be contacting several firms at once. If your line goes to voicemail at 7 p.m. or your intake team is on another call, that person often keeps dialing.
An AI receptionist built for intake answers immediately, any time, in English or Spanish, and moves the caller to a concrete next step: a booked consultation, a document request or a callback from an attorney.
What does an AI intake agent do?
Answers and identifies the matter type
The agent greets the caller with your firm’s name, explains that it’s an AI assistant and asks how it can help. Within a few questions it can tell an existing client calling about their case (route to the right team) from a new inquiry (start intake) and from calls that aren’t yours at all (vendors, wrong numbers, solicitations).
Screens with your criteria
You define what makes an inquiry a fit: practice areas you handle, states and counties where you practice, matter types you decline, minimum criteria your attorneys set. The agent asks the questions in your intake script and tags the inquiry. It doesn’t decide on its own that someone "has a case." If a matter looks time-sensitive, for example the caller mentions a court date or an incident from a while ago, the agent flags it for prompt attorney review instead of commenting on deadlines.
Collects conflict-check information first
Before the caller tells the whole story, the agent collects the names of the people and companies involved, so your team can run a conflict check. Keeping the initial details limited until conflicts are cleared is a practice many firms already follow with human intake staff; the AI can apply it consistently.
Books the consultation
The agent offers open slots from the attorney’s calendar based on practice area, location and consult type (phone, video or in person), books it and sends a confirmation with any forms or documents to bring. If your firm charges for consultations, it states the fee exactly as you wrote it.
Hands everything to your intake system
A structured summary, the transcript, the tags and the conflict-check names go into your intake CRM or practice management software, such as Clio Grow or Lawmatics, or whatever your firm uses. Here’s an illustrative example of an intake email flowing into a firm’s system:
Illustrative example Fictional business, people and data. Not a real client, conversation or result.
Where intake ends and legal advice begins
This line needs to be written into the agent’s instructions and tested, because callers will ask: "Do I have a case?" "Should I sign this?" "What will I get?"
Intake vs legal advice
| Caller asks | What the AI does | What it never does |
|---|---|---|
| "Do I have a case?" | Explains that an attorney will review the facts and offers a consultation | Says yes, no or "probably" |
| "How much is my case worth?" | Explains that only an attorney can evaluate that after a review | Gives a number or a range |
| "Is there a deadline?" | Flags the matter as time-sensitive for prompt attorney review | Comments on statutes of limitations |
| "Are you a lawyer?" | Says it is an AI assistant for the firm | Implies it is a person or an attorney |
| "Will you take my case?" | Explains the firm’s next step and timing | Promises representation |
What the AI does
- "Do I have a case?"
- Explains that an attorney will review the facts and offers a consultation
- "How much is my case worth?"
- Explains that only an attorney can evaluate that after a review
- "Is there a deadline?"
- Flags the matter as time-sensitive for prompt attorney review
- "Are you a lawyer?"
- Says it is an AI assistant for the firm
- "Will you take my case?"
- Explains the firm’s next step and timing
What it never does
- "Do I have a case?"
- Says yes, no or "probably"
- "How much is my case worth?"
- Gives a number or a range
- "Is there a deadline?"
- Comments on statutes of limitations
- "Are you a lawyer?"
- Implies it is a person or an attorney
- "Will you take my case?"
- Promises representation
Ethics points to review with your state bar’s rules
We’re not your ethics counsel, and nothing here is legal or ethics advice. What we can do is point you to the topics that come up whenever a firm uses AI for intake, so you can review them with your state’s rules and your counsel. The ABA Model Rules are the common reference, but each state adopts its own version and interprets it through its own opinions.
- Confidentiality (Model Rule 1.6). Intake calls contain confidential information. Review how the AI vendor stores, uses and retains call data, and whether it’s used to train models.
- Duties to prospective clients (Model Rule 1.18). Information from people who never become clients can still be protected and can create conflicts. Ask how your intake design limits what’s collected before a conflict check.
- Supervision of nonlawyer assistance (Model Rule 5.3). Firms are responsible for the work of assistants they use. Review how your firm supervises and audits the AI’s conversations.
- Unauthorized practice of law (Model Rule 5.5). The AI must not give legal advice; your script and testing should prove that.
- Communications and solicitation (Model Rules 7.1 and 7.3). What the AI says about your firm is a communication about your services. If you plan outbound calls or texts to prospects, review the solicitation rules first.
- Competence (Model Rule 1.1). Lawyers are expected to understand the benefits and risks of the technology they use.
For AI specifically, the ABA issued Formal Opinion 512 on generative AI tools in July 2024, and several state bars have published their own guidance, including the State Bar of California (November 2023) and The Florida Bar (Ethics Opinion 24-1, January 2024). Your state may have its own opinion or none yet. Check your state bar’s ethics opinions page, and when in doubt, use its ethics hotline if it offers one.
Two more items that aren’t bar rules but come up in every project: call recording consent, which varies by state and may require all parties to agree, and outbound calling and texting rules under the TCPA if you plan to follow up automatically. Our TCPA overview for AI calls explains the basics to discuss with counsel.
How to set it up
Launching AI intake at a law firm
-
Step 1
Write the intake script with your attorneys
practice areas, screening questions, declined matters, approved disclaimers
-
Step 2
Define the conflict-check fields
which names the AI collects first and where they go
-
Step 3
Connect calendars and intake software
consult types, attorney availability, CRM fields
-
Step 4
Set escalation rules
urgent matters, existing clients, distressed callers and anything outside the script
-
Step 5
Test the advice boundary
role-play the questions callers really ask and check every answer
-
Step 6
Launch after hours first
then add daytime overflow and web chat once reviews look right
-
Step 7
Review conversations weekly
an attorney or intake lead audits a sample and updates the script
Phone, chat or both?
Phone is where most urgent legal inquiries start, but many people fill out a web form or open a chat at night. The same intake logic can run across channels: a voice agent on the phone line and an AI chatbot on the website, both writing into the same intake system. If you’re comparing AI with a human legal answering service, our article on AI receptionists vs answering services lays out the trade-offs.
How we build intake for law firms
Appyweb is an AI automation company headquartered in Austin, Texas, building software since 2012. For law firms we design the intake flow around your screening criteria and your conflict process, connect it to the calendars and intake software you already use, and make the advice boundary a tested part of the system, not an afterthought. Follow-up steps, such as sending engagement documents after an attorney accepts a matter, can be handled by workflow automation.
If new inquiries are slipping through after hours or while your team is on other calls, see how our AI receptionist works, or tell us how intake runs at your firm today.