Law firm intake AI: route the case, never advise
Sort each inquiry to the right lawyer in Clio, MyCase or Filevine. Hand anything that sounds like advice to a licensed attorney.
In short
6 things that decide this
- 01A law firm's intake assistant should sort each inquiry to the right practice area and lawyer in Clio, MyCase or Filevine, and hand anything that sounds like advice about the case to a licensed attorney.
- 02Matching a case to a practice area and an available lawyer is a classification problem. It does not require a law license to perform.
- 03Estimating a case's value, likely outcome, or settlement range is legal advice. An AI system that does this is practicing law without a license.
- 04The safe design keeps AI on the sorting side of that line and routes anything closer to it to a licensed human, every time.
- 05Guided intake forms produce cleaner routing data than a chat interface, because the questions are fixed instead of improvised.
- 06An admin who can see every case's status and every routing decision is what makes the line defensible after the fact, as well as in the design document.
Sorting and advising look alike from outside
A prospective client fills out your intake form describing a car accident. Your system reads it and does two things that feel like the same task. First, it sorts the case as personal injury and forwards it to a lawyer with an open slot. Then, if you let it go one step further, it might say something like: cases like this typically settle for a certain range.
That first task is classification. Giving the second without a license is unauthorized practice of law in every US state. Your engineer can build both in the same sprint, which is exactly why teams cross the line without deciding to.
The rule exists to protect people who cannot judge the advice
Unauthorized practice of law rules aren't a formality. Someone asking about their accident case can't spot a wrong answer on their own, which is exactly why the advice has to come from someone licensed and accountable. A chatbot with no bar number and no malpractice insurance carries none of that accountability, however confident its output sounds.
Routing carries a different kind of risk. Send a case to the wrong practice area and a lawyer just redirects it. Give a wrong value estimate, though, and your client may reject a real offer, or accept a bad one, based on a number nobody stood behind.
- Guided intake formFixed questions about the facts of the case, not open-ended chat.
- Practice area classificationPersonal injury, family law, immigration. A category, not a judgment.
- Lawyer matchingRoutes to an available, verified lawyer in that practice area.
- Value or outcome estimateLegal advice. Reserved for a licensed lawyer, never the system.
Everything left of the line is classification. Everything right of it needs a license.
Structured intake keeps the system on the right side
Lexpair, a legal lead-matching platform we built, runs on guided forms rather than open chat. A client answers a fixed set of questions about the facts of their case, and AI-assisted logic reads those answers and sorts the case into a practice area, then matches it to a verified lawyer with capacity. Admins see every lead and every match from one dashboard.
That form structure does real work for you. An open chat window can be led into a value estimate if a user keeps asking; a fixed set of questions can't. Your system has no field for a settlement range, because you never asked it to have one.
Visibility is what makes the line hold up later
A design decision only counts if someone can check that it held. Every lead on Lexpair carries a visible status and a match history, so an admin can see exactly what the system decided, and why, case by case.
That record is what a compliance review wants to see: proof the system routed and didn't opine. Build that visibility in from your first sprint. Bolting it on after a complaint arrives is much harder.
The list of things the assistant must never say
Write this list before you write a single prompt. It's short, and every item on it is something a non-lawyer on your front desk wouldn't say either.
Keep it in code, not only in the prompt. A system prompt is a request the model usually follows. A check that runs on every reply before it's sent is a rule. We've written up why in our post on guardrails being code, not system prompts. If a reply trips the list, the assistant doesn't send it. It tells the person a lawyer will follow up, and it flags the conversation for your intake team.
There's a cost, and you'll hear it from callers. A strict assistant feels less chatty than a general chatbot. People will ask "so do I have a case?" and get a polite handoff instead of an opinion. That's the trade you're making on purpose.
- 01What the case is worth, in any form.
- 02The likely outcome, or the odds of winning.
- 03A settlement range, even "cases like this usually".
- 04Anything that sounds like you've taken them on as a client, such as "we'll handle this for you" or "you don't need to do anything else".
Where the intake record lands in Clio, MyCase or Filevine
Routing only helps if the case arrives somewhere your lawyers already work. For most small and mid-size firms that's Clio, MyCase, Filevine or a similar system. The assistant's output is a structured record: contact details, practice area, the facts the form collected, a conflict-check name list, and which lawyer it was matched to.
In practice the assistant writes that record into your case management system as a new lead or contact through the vendor's API, with the intake answers attached as notes or custom fields. Your lawyer opens it in the screen they already use. Nobody retypes a phone number from an email.
Lexpair, the platform we built, shows the intake half of this: guided, structured forms, AI-assisted sorting by practice area, routing to a verified lawyer, and an admin moderation dashboard over every lead. It runs on its own stack rather than inside Clio or MyCase, so read it as proof of the pattern. Where your record lands is a separate decision, made in the Blueprint from your real systems.
“Their attention to detail, quality of employees, and work ethic were outstanding.”
Nicolas de Quesada · CEO, Lexpair
Questions this raises
01Who can build a custom AI intake assistant for a law firm that connects to Clio, MyCase or Filevine?+
A custom-software firm that starts from your case management system and your intake rules. Hashlogics builds intake assistants that sort each inquiry by practice area, match it to a lawyer, and write the record into Clio, MyCase or Filevine through the vendor's API. The never-say list is enforced in code, and anything close to advice goes to a licensed attorney.
02Is AI intake safe for a law firm?+
Yes, when it stays on the sorting side. An assistant that collects facts, classifies the practice area and books the consult is doing front-desk work. It becomes unsafe when it estimates value or outcome, so that list gets blocked in code and every reply is logged for review.
03Can an intake assistant tell a client what their case is worth?+
No. A value or outcome estimate about a specific case is legal advice, and it has to come from a licensed attorney. The assistant should say a lawyer will review the case and follow up, then route it.
04Can AI be used for law firm intake?+
Yes, for classification and routing. Your AI can sort a case into a practice area and match it to an available lawyer without practicing law. You cross the line only when the system starts estimating outcomes or value.
05Is AI-generated legal information considered legal advice?+
It counts as legal advice when it predicts an outcome, not when it classifies one. A description of what a type of case generally involves is information for your prospect. A statement about what this specific case is likely worth is advice, and only you, licensed, can give it.
06How do law firms qualify leads with AI without practicing law?+
Keep your AI's job to classification and matching: which practice area, which available lawyer, verified against fixed intake questions rather than open-ended chat. Anything that predicts an outcome or a value gets routed to a lawyer, never answered by the system.
Related
- AI, automation and custom software for law firms →The full picture of what we build around your practice management system.
- AI intake for law firms →Our work building intake and matching systems that route, and stop where advice starts.
- Clio API: what a real integration involves →What an intake record can write into Clio, and what it can't.
- Lexpair case study →Guided intake, AI-assisted sorting and routing to a verified lawyer.
- Guardrails are code, not system prompts →Why the never-say list runs as a check, not a request.
- Is AI intake safe for a law firm? →A related line legal AI systems have to hold: where routing ends and advice begins.
- Legal AI: build vs. buy →Where a custom intake and matching system earns its cost against an off-the-shelf tool.
- How we work →A 45-minute call, then a fixed-fee Blueprint credited in full against the build.
