A prospective client calls about a property dispute. Someone, usually a paralegal and sometimes an associate, steps away from a filing to take the call and spends the first ten minutes asking the background questions the firm asks on every new matter: what happened, who's involved, when, whether there's a deadline, and how to reach them.
Only after that does anyone know whether it's even a case the firm takes. When the paralegal is in a deposition prep or on another call, the prospective client gets voicemail, and many of them simply call the next firm on their list.
An AI receptionist for law firms exists to handle those first ten minutes consistently, at any hour, without pulling someone off billable work. It answers calls and website chats, collects structured intake facts, books consultations for matters that fit, and routes urgent or sensitive calls to a person immediately.
What it never does is give legal advice, and for most partners that boundary is the first thing to settle. Below we show where it sits, how to write it into the assistant's rules, and how to test it, with invented intake scenarios along the way.
What is an AI receptionist for law firms?
It's software that answers the firm's phone and website chat and runs the first stage of client intake. It gathers the facts your team needs to decide whether to take a matter, then hands those facts to a person.
Intake is data collection, not legal advice, and the distinction matters enough to state plainly. A well-configured assistant doesn't assess the merits of a case, offer an opinion on what someone should do, or answer legal questions. It collects matter type, parties involved, key dates, urgency, and contact details.
The boundary is set in the assistant's rules: gather facts, never advise, never comment on a case's merits, and send every legal question to a person. Test it deliberately before launch by asking it legal questions and confirming it declines.
For most firms, intake starts on the phone, so the assistant answers the firm's Twilio number. The same intake can run in the chat box on your website, and with Booking by text on, a prospective client can book, move, or cancel a consultation by text. Other texts wait for staff, who can ask for a suggested reply and send it themselves. It fits solo practitioners, small firms, and mid-size practices with several practice areas, especially where a paralegal or receptionist currently handles intake alongside other work. The law firm solutions page summarizes the setup.
How AI legal intake works
Step 1: The call is answered immediately
Every call is picked up, including evenings and weekends. Firms that want the receptionist to keep answering in office hours can route calls to the front desk first, with Aiva Chat covering whatever goes unanswered.
Step 2: It identifies the type of matter
The caller describes their situation in their own words. A caller who says their landlord is trying to evict them without proper notice is routed into your property or tenancy intake, not a generic form.
Step 3: It asks the practice-area questions
For that eviction call, it asks what a paralegal would: when the notice was given, what it said, whether there's a written lease, how long they've been a tenant, and whether there's a court date or response deadline. It doesn't say whether the eviction is lawful.
Step 4: It captures conflict-check details
It records the caller's name, the other party's name and details, and the nature of the dispute, accurately and consistently. Your team runs the actual conflict check.
Step 5: It books, routes, or escalates
If the matter looks like a fit, Aiva Chat books a consultation and the intake summary goes to HubSpot or your case pipeline. It only offers times inside your consultation hours that don't clash with existing Aiva Chat bookings or busy times on the connected Google Calendar, so meetings a lawyer already has on that calendar are respected. Add each lawyer as a staff resource with their own hours and two prospective clients can book the same hour with different lawyers. Where a matter needs a conflict check first, per-service approval turns the booking into a request that holds the slot until someone accepts it, and it expires after 24 hours by default if nobody does. Callers in distress, facing an urgent deadline, or asking for a person are transferred straight to the lawyer or department you've set up.
Benefits for a law firm
Paralegals get back to substantive work
A meaningful share of a paralegal's day can go to qualifying calls that turn out not to fit the firm's practice areas. When the first ten minutes are handled consistently, paralegals open a structured summary instead of starting with "so, tell me what happened."
Fewer prospective clients lost to phone tag
A prospective client who can't get a straight answer quickly often calls the next firm. The matter is gone before anyone assessed it, and the case may have been a strong one; the only problem was a two-day callback.
Consistent intake on every matter
Human intake drifts with workload. On a quiet afternoon, every question gets asked; on a busy day, someone grabs a name and number. The assistant asks the same practice-area questions every time, which also makes conflict checks faster.
Intake in the caller's language
Stress pushes people back toward their most fluent language. The assistant hears which language the caller is using and runs intake in it, from 13 supported languages, so a caller describing a dispute in Gujarati or Hindi gets the same structured intake as an English speaker.
Visibility for the managing partner
Most small firms have no clean picture of how many enquiries arrive, which practice areas they concern, or how many come in after hours. Topic tagging, lead and booking counts, and plain-language questions across calls make that visible, which helps when deciding where to add capacity.
Law firm intake use cases
The eviction call at 7 PM
Problem: A tenant calls after hours, anxious about an eviction notice with a short response deadline. How Aiva Chat handles it: It runs property intake, captures the notice date and deadline, and because a deadline is close, transfers to the on-call number your rules specify. Outcome: A lawyer speaks to the caller that evening with the facts already gathered.
The family law enquiry that needs care first
Problem: A caller begins describing a separation and mentions feeling unsafe at home. How Aiva Chat handles it: Your rules treat safety concerns as an immediate handoff, so the call is transferred to a person rather than continuing with routine questions. Outcome: The caller reaches a human straight away, which is the only appropriate response.
The contract question that isn't a fit
Problem: A business owner calls about an intellectual property licence, a practice area the firm doesn't handle. How Aiva Chat handles it: It explains the firm doesn't take that type of matter, captures contact details if your rules ask for them, and doesn't offer an opinion. Outcome: No paralegal time spent, and the caller gets a clear answer instead of a callback that never comes.
The weekday overflow
Problem: Three new-matter calls arrive while the only intake paralegal is on another call. How Aiva Chat handles it: Desk phones ring first, then Aiva Chat picks up each unanswered call, runs intake, and books two consultations. Three calls at once fits the Growth plan's limit of five simultaneous calls; Starter handles two. Outcome: The paralegal finishes their call to find two booked consultations with summaries attached.
AI receptionist vs. a paralegal or a legal answering service
| Intake paralegal | Legal answering service | AI receptionist (Aiva Chat) | |
|---|---|---|---|
| Availability | Office hours, one call at a time | Extended hours | Every call, any hour |
| Practice-area intake depth | Excellent | Usually a short script | Follows the questions you configure |
| Judgment on urgency and sensitivity | Excellent | Limited | Follows your rules, transfers to people |
| Legal advice | Can't give it either | No | No |
| Languages | Whatever staff speak | Varies | 13, auto-detected |
| How it's billed | Salary and benefits | Per call or per minute | Monthly plan; the firm's own Twilio and OpenAI accounts carry call and AI usage |
An experienced intake paralegal is better than any software at reading a caller, sensing what isn't being said, and judging which matters deserve a partner's immediate attention. The constraint is capacity: one call at a time, office hours, and other work competing for the same hours.
Legal answering services extend hours, but most run a short script and take a message, so the real intake still happens on the callback. The AI receptionist runs fuller, practice-specific intake at any hour and passes anything that needs judgment to a person. Most firms use it to handle volume and after-hours calls, keeping their people on the conversations that need them. Our explainer on how AI hands calls to humans covers the handoff mechanics in detail.
How to set up an AI receptionist for your firm
The dashboard's six setup steps handle the mechanics. For a firm, the real work is writing the boundary and the intake questions, so those come first.
Step 1: Define the intake boundary
Write down what the assistant may do (collect facts, book consultations, explain practice areas and fees for initial consultations) and what it must never do (advise, assess merits, promise outcomes). These become rules.
Step 2: Write intake questions per practice area
For each practice area, list the questions your best paralegal asks, in order. Include what counts as urgent, such as an imminent court date, and what should trigger an immediate handoff.
Step 3: Teach Aiva Chat about the firm
Add practice areas, the matters you don't take, consultation fees and format, office address, and hours. Import pages from your website or upload a PDF.
Step 4: Connect calendar, CRM, and transfers
Move or forward the firm's intake line to a Twilio number, add your OpenAI key, link the consultations calendar in Google Calendar, and connect HubSpot or set up webhooks for your case pipeline; the integrations page lists the options. Set up transfer departments with trigger phrases for urgent matters.
Step 5: Test the boundary hard
Ask it legal questions directly: "Will I win?" "Is this eviction legal?" Confirm it declines every time. Test an urgent call, a distressed caller, and a matter you don't take.
Step 6: Review intake quality weekly
Read a sample of transcripts, check that conflict details are complete, and refine questions where intake summaries come out thin.
Common mistakes firms make with AI intake
Leaving the advice boundary implicit
If the rules don't explicitly forbid advice, a helpful assistant may drift into explaining what a law "usually" means. Write the boundary as explicit rules, test it with direct legal questions, and recheck after any change to the assistant's instructions.
One generic intake script for every practice area
A family law matter and a commercial dispute need different facts. A generic form produces summaries that are too thin to act on, which pushes the real intake back onto the callback.
No route for urgent matters
Without trigger phrases and an on-call number, an urgent matter at 9 PM gets the same treatment as a routine enquiry. Decide in advance what counts as urgent and who receives those transfers.
Skipping the data-handling conversation
Prospective clients share sensitive information before any engagement exists. Set a retention period that matches your firm's policies, and read the security page as part of your diligence. The FTC's guide on protecting personal information is a sensible baseline for thinking about what to keep.
Not reading the transcripts
Intake quality only improves if someone reads what callers said and where the questions fell short.
Best practices for law firm AI intake
- Write explicit "never" rules for advice, merits, and outcome predictions.
- Build separate intake paths for each practice area you take.
- Capture opposing party details on every matter for conflict checks.
- Set trigger phrases for urgent and sensitive situations, with a real person on the other end.
- Keep consultation fees and process clear so callers know what to expect.
- Set retention deliberately rather than keeping everything forever. Our post on AI phone agent data privacy covers the trade-offs.
- Review a sample of intakes weekly in the first month.
Conclusion
For a small firm, an AI receptionist is worth it if new-matter calls regularly reach voicemail or wait a day for a callback. Intake is mostly fact-gathering, and handing that first stage to a consistent assistant frees paralegal hours without changing who makes legal judgments.
The make-or-break decision is the advice boundary. Write it as explicit rules, test it by asking the assistant whether a caller will win, and recheck it after every change to its instructions. Weigh data handling honestly too: Aiva Chat offers retention controls and transcript scrubbing but doesn't hold formal compliance certifications today, and some firms will need that.
A practical first step is to count last month's new-matter calls that went to voicemail or waited more than a day for a callback.
Ready to take intake off your paralegal's desk?
Build one practice area first, say tenancy, and run a dozen test calls that try to pull legal advice out of it before any real client hears it.
The 14-day trial needs no card. After it, the firm pays $49, $99, or $199 a month depending on numbers and simultaneous calls, while call minutes, texts, and AI usage are billed to the firm by Twilio and OpenAI at their own rates. Call +1 (970) 972-0974 to hear an intake call handled.
Start your free trial, or book a demo to map your practice areas with us.
Posts from the people building Aiva Chat in Rajkot — how the product works, what we have learned building it, and guides for getting the most out of an AI receptionist.