AI intake for personal injury law firms: a practical guide

AI intake for personal injury law firms means an AI system that answers every call, web chat and social media message the moment it arrives, asks your qualification questions in a natural conversation, and books qualified prospects straight into a consult. It hands urgent or unclear matters to your intake specialist or an attorney, and it passes a complete, de-duplicated record into your case management system. Done well, attorneys stop spending time on calls that were never cases, and no injured person waits until Monday morning for an answer.

Below: the questions, escalation rules, metrics and a rollout plan, with a real Florida PI build as the worked example.

What does AI intake for a personal injury law firm actually do?

AI intake does five jobs, each one something your team does by hand today, usually only during office hours.

1. Answers instantly on every channel. A voice agent picks up the phone. Chat agents respond on your website and in Instagram, Facebook Messenger and TikTok DMs. All of them run the same intake logic.
2. Qualifies through conversation. It collects the facts your intake team needs, in plain language, one question at a time.
3. Applies your criteria consistently. The rules you set (practice area, jurisdiction, representation status and so on) are checked the same way at 2 p.m. and 2 a.m.
4. Books the next step. Qualified prospects get a consult slot, a place to send evidence and reminders before the appointment.
5. Hands off cleanly. Your team gets an alert and a summary, and the lead lands in your CRM or case management system once, not three times.

What it does not do: give legal advice, tell anyone they have a case, quote a settlement value or sign anyone up. Those stay with licensed people.

Phone, web chat and social DMs

Each channel behaves differently, so the agent has to adapt while keeping one set of rules underneath.

• Phone. The most urgent inquiries still come by phone. An AI voice agent needs to sound calm, handle interruptions, confirm spelling of names and numbers, and know when to transfer live.
• Website chat. Visitors are often comparing firms. A chat agent should start the intake conversation rather than push a contact form.
• Instagram, Facebook and TikTok DMs. If your firm creates content, this is where a lot of first contact happens, often within hours of an accident. Messages are short and informal. A good AI chat agent asks for a phone number early so the conversation can continue if the person goes quiet.

How do qualification criteria work in PI law firm lead intake?

Qualification is where most of the value sits, and it is where the firm, not the vendor, has to make the decisions.

Most PI firms screen on the same core facts. The table below shows what to ask, why it matters and how the system typically treats the answer.

Criterion	What the AI asks	Why it matters	Typical handling
Incident type	“Can you tell me what happened?” then a follow-up to classify it (car crash, truck, slip and fall, dog bite, medical, workplace)	Confirms the matter is in your practice area	Outside your practice area: polite decline, optional referral message
Date of incident	“When did this happen?”	The statute of limitations window and how fresh the evidence is	Flag any date close to your firm’s limitations cutoff for urgent human review; never auto-decline on date alone
Location	“Where did the accident happen, and where do you live now?”	Jurisdiction and whether your attorneys are licensed there	Out-of-state: route to your referral rule
Injury	“Were you hurt? Have you seen a doctor or been to the ER?”	Whether there is a physical injury and treatment to document	No injury reported: route per your property-damage policy
Liability	“Who do you think was at fault? Was a police report made?”	An early read on fault and available evidence	Captured for the attorney; rarely a hard gate
Insurance	“Do you know if the other driver had insurance? Do you have your own policy details?”	Recovery potential	Captured as information, not a gate
Prior representation	“Are you already working with a lawyer on this?”	Avoids contacting someone who is represented	Already represented: end intake politely
Contact details	Full name and a phone number, confirmed back	Without these, nobody can follow up	Required before booking

Hard gates versus information to collect

Separate the questions that decide whether a lead moves forward (hard gates) from the ones that just enrich the file. Keep hard gates few and objective: residency, injury, representation status, valid contact details. Fault and insurance are judgment calls, so capture them and let a person weigh them.

On the Awais AI Guy Show, Kweku Darfoor, managing attorney of Darfoor Law Firm, made the case that AI should never reject a case without an attorney reviewing it. Whatever your rules, keep a record of every screen-out so a person can look at the declines each week. That is how you catch a rule that is too strict.

Sample opening script

“Hi, you’ve reached the intake line for [Firm]. I’m an AI assistant and I can help get your information to our team right away. I’m sorry you’re dealing with this. Can you tell me a little about what happened?”

Note the disclosure that the person is talking to an AI, the empathy before the questions, and the open question first. Firms should check their state bar’s rules on advertising, solicitation and supervision of non-lawyer assistance before finalizing any script, including how the assistant identifies itself.

When should AI intake escalate to an intake specialist or attorney?

Escalation rules are what make AI intake safe. Define them up front and test them before launch. Common triggers:

• Catastrophic injury or death. Wrongful death, brain injury, spinal injury, amputation or a hospitalized caller. These go to a person immediately.
• A limitations deadline is close. Anything near your cutoff gets same-day human review.
• Distress or confusion. If the person is crying, in shock or cannot follow the questions, stop qualifying and transfer.
• Specific legal questions. “Should I talk to the insurance adjuster?” is a question for a lawyer. The AI says so and offers to connect them.
• The caller asks for a person. Always honor it.
• Edge cases your rules don’t cover. Default to a human, not a decline.

The best escalation is a warm transfer: the AI briefs the intake specialist or attorney on who is calling and what happened, then connects the call. Nobody asks the injured person to repeat their story. After hours, escalation becomes an alert to whoever is on call plus a booked callback.

How does AI intake book the consult and hand off to your CRM?

Qualification is wasted if the prospect leaves the conversation without a next step. The moment a lead clears your gates, the system should do the following on its own:

1. Offer real consult slots from your calendar, or send a booking link pre-filled with what the prospect already said.
2. Send a place to upload evidence (photos, police report, medical records, insurance cards), filed under that lead.
3. Alert your team in real time (Slack, Teams, SMS or email) with a short summary.
4. Create or update the record in your CRM or case management system, with the transcript or call recording attached.
5. Send reminders before the consult.

De-duplication matters more than you think

The same person often reaches out twice: a DM at night, then a call in the morning. Without de-duplication, two staff members chase the same lead and your pipeline numbers are wrong. Match on phone number at minimum, plus the chat session, and merge rather than create.

Missed-call recovery

Even with a voice agent, some calls drop or ring through during an outage. Set up an instant text to the caller acknowledging the call and an alert to staff.

A worked example: AI intake at Darfoor Law Firm

Darfoor Law Firm is a Fort Lauderdale, Florida personal injury practice handling accident, malpractice and wrongful death matters. Its social media presence on Instagram, Facebook and TikTok produced steady inbound demand, but calls and DMs after hours waited until the next business day. Intake was manual, staff re-keyed details by hand, records duplicated, and missed calls had no follow-up.

Over four weeks we built an agentic AI intake system:

• AI chat agents on two Instagram accounts, Facebook Messenger and TikTok, plus a voice AI receptionist on the phone line, all running the same logic.
• An eight-step qualification conversation collecting name, location, representation status, injury details, accident type and available evidence.
• Five gates checked automatically before anything reaches an attorney: Florida residency, confirmed physical injury, unrepresented status, a valid phone number and a complete name.
• For qualified leads: a pre-filled booking link, a dedicated evidence folder, a real-time Slack alert and reminders in the 24 hours before the consult.
• Urgent calls go live to Mr. Darfoor after the voice agent briefs him on the situation.
• Every missed call triggers two texts within seconds, one to the prospect and one to staff, and every lead is de-duplicated by phone number and session.

The results, as recorded in the Darfoor Law Firm case study:

• 4 → 1: Instagram (two accounts), Facebook Messenger, TikTok and phone unified into one intake pipeline.
• 100% of inbound leads captured and answered, regardless of time of day or staff availability.
• 0 duplicate leads entering the pipeline, thanks to de-duplication by phone number and session.
• 5 qualification gates applied automatically before anything reaches an attorney.
• 24 hrs of automated reminders before every booked consultation.

What should you measure after launching personal injury intake automation?

Measure before you launch, so you have a baseline. Pull at least a month of call logs, DM history and CRM records. Then track these weekly.

Metric	How to calculate it	What it tells you
Speed to first response	Time from first inbound touch to first reply, by channel	Whether anyone is left waiting
After-hours captures	Leads that arrived outside office hours and completed intake	Demand you were previously losing overnight and on weekends
Qualified rate	Leads passing all gates ÷ leads that started intake	Lead quality by channel, and whether your gates are too tight or loose
Completion rate	Conversations that reached the end ÷ conversations started	Where people drop off in the question flow
Consult booking rate	Consults booked ÷ qualified leads	Whether the handoff to the calendar works
Consult show rate	Consults attended ÷ consults booked	Whether reminders and follow-up are doing their job
Escalation rate	Conversations handed to a person ÷ total	Whether escalation rules fire as intended
Declines reviewed and overturned	Screen-outs a person changed to “pursue”	Rules that need loosening

Read transcripts every week in the first month.

What parts of PI intake should stay human?

AI handles the volume and the repetition. People keep the judgment and the relationship. A sensible split:

• AI handles: first response on every channel, fact collection, applying hard gates, booking, reminders, evidence collection, CRM entry, missed-call follow-up and the after-hours queue.
• People handle: anything that sounds like legal advice, case evaluation, borderline or declined matters, catastrophic injury and wrongful death conversations, fee discussions, engagement agreements and the consult itself.

Your intake specialists don’t disappear. They stop typing details off voicemails and start spending time on the calls that need a person. Firms should also check their state bar’s rules on supervision of non-lawyer assistance, since AI intake operates on the firm’s behalf.

How do you roll out AI legal intake without losing cases?

A phased rollout lets you prove the system on low-risk traffic before it handles everything. A typical sequence:

1. Map current intake (week 1). List every channel, who answers it and when, what questions get asked, where records go and where leads fall through. Write down your hard gates and escalation triggers. If you want a structured way to do this, our free Case-Flow Checkup for PI firms walks through intake, records, updates and handoffs.
2. Build and test internally (weeks 1–3). Configure the voice and chat agents, connect your calendar and CRM, and run test calls and DMs from your own staff: easy cases, out-of-state callers, represented claimants, distressed callers, wrongful death scenarios.
3. Go live after hours first. Let the AI take nights and weekends while your team keeps daytime. This is where the gain is largest and the risk to existing processes lowest.
4. Add social DMs and web chat. Turn on one channel at a time and check transcripts daily for the first week of each.
5. Extend to daytime overflow. The AI answers when your team is busy, with warm transfer when someone frees up.
6. Review weekly and tune. Adjust questions where people drop off, tighten or loosen gates based on reviewed declines, and refine escalation triggers.

For a single system like this, our published pricing starts with a build sprint from $4,500, typically two to four weeks. The Darfoor build took four weeks across five entry points (two Instagram accounts, Facebook, TikTok and phone).

Where to start

Start with three numbers you can pull this week: how many inquiries arrived after hours last month, how long the average after-hours lead waited for a reply, and how many records in your CRM are duplicates. If those numbers are uncomfortable, the case for AI intake is already made.

Then write one page: your hard gates, your escalation triggers and what a qualified lead should receive the moment it qualifies. That page is the specification any build, ours or anyone else’s, should follow. For more on how we approach legal intake, see our law firm page.