BUILT FOR SOLO AND SMALL FIRM ATTORNEYS

AI for the work behind the practice of law

Intake calls stack up while you draft. Document review eats the day before the brief gets touched. AI handles the first-pass work so attorney time goes where it has to: judgment, strategy, and client counsel that respects privilege.

6 hrs/wk
Back from intake triage
3x
First-pass doc review throughput
50% faster
Demand letter and brief drafts
4 hrs/wk
Billable time reclaimed

The short answer

Attorneys can use AI without violating ABA Model Rules or facing Mata-style sanctions when AI is treated as junior-associate work, not partner work. The rule is simple: every citation gets verified against the actual case, client and privileged information stays inside confidential, firm-deployed environments, and the attorney still practices law. AI drafts. The attorney signs.

Why solo and small firm attorneys are using AI right now

AI handles the admin layer and first-draft work. Final attorney review stays exactly where it has always been: on the attorney.

Client intake at scale

Triage incoming matters before your first call of the day. Pull facts, flag conflicts to check, classify the practice area, and surface the questions worth your time. The unqualified leads stop eating your morning.

Document review, faster first pass

Run a structured first read across discovery productions, contracts, or filings. AI flags the clauses, the dates, the parties, and the issues that warrant attorney attention. You review what matters instead of every page.

Contracts drafted with attorney review built in

Generate a working draft from your firm's clause library and the matter facts. Every output ships as a draft, marked for attorney redline. The blank page is gone. The attorney judgment is not.

Research summaries that point to source citations

Get a structured summary of a research question with source citations you verify in Westlaw or Lexis. Citations are leads, not authority. Every cited case gets read by a human before it lands in a filing.

Demand letters and briefs as drafts only

First drafts of demand letters, motion shells, and brief sections from your facts and a target outline. The structure is there in minutes. The attorney work is the editing, the strategy, and the final sign-off.

Client comms volume handled

Status update emails, billing explanations, and routine client questions drafted in your firm voice. You edit and send instead of starting from scratch on every message. Response times drop. Client satisfaction climbs.

AI in your law practice, specifically

Think of AI as junior associate work, not partner work. Here is the team you have access to. Every output gets attorney review before it leaves the firm.

AI as an Intake Triager

Sort new inquiries by practice area, urgency, and conflict-check needs. Pull the facts worth your attention before you take the call. The work that should not have made it past intake never does.

Looks like
You are an intake assistant for a small estate planning firm. Read this inquiry and produce: practice area classification, three conflict-check names to run, a five-bullet fact summary, the top three clarifying questions for the attorney to ask on the first call. Flag anything that suggests urgency. Output is for attorney review before any client contact.

AI as a Document Reviewer

First-pass review across contracts, discovery, or filings. Flag clauses, dates, and issues that need attorney eyes. The reviewer never replaces attorney judgment, only narrows where it goes.

Looks like
You are reviewing a commercial lease as a first-pass reviewer. Identify and flag: assignment and subletting restrictions, renewal terms, indemnification scope, default and cure periods, holdover rent, and any unusual operating-expense pass-throughs. Output a flag list with section references for attorney review. Do not draft client advice.

AI as a Contract Drafter

Generate a working contract draft from your clause library and the matter facts. The output ships marked DRAFT. Attorney redline and final approval are required before any client sees it.

Looks like
Draft an independent contractor agreement using these facts: parties, scope, payment terms, IP assignment, jurisdiction, term. Use plain English where possible. Mark every variable in brackets for attorney review. Label the entire output DRAFT - FOR ATTORNEY REVIEW. Do not finalize or send.

AI as a Research Summarizer

Summarize a research question and point to source citations as leads. Every citation must be pulled and read in Westlaw or Lexis before it lands in a filing. AI-generated citations are starting points, never authority.

Looks like
Summarize the law in [jurisdiction] on enforceability of non-compete agreements for sales employees. Provide a one-paragraph summary, three case citations as research leads, and the statutory framework if any. The attorney will verify every citation in Westlaw before relying on it. Do not invent or paraphrase holdings.

AI as a Demand Letter and Brief First-Drafter

Produce a structured first draft of a demand letter or brief section from facts and a target outline. The strategy, the tone calibration, and the final sign-off stay with the attorney.

Looks like
Draft a demand letter to a former employer alleging unpaid wages under [state] law. Use this fact pattern and outline. Tone: firm but professional. Mark every legal conclusion in brackets for attorney verification. Label the output DRAFT - FOR ATTORNEY REVIEW. Include a request for response within 14 days.

AI as a Client Communicator

Draft routine client emails, status updates, and billing explanations in your firm voice. Attorney reviews and sends. The volume of admin-layer client comms stops eating partner time.

Looks like
Draft a status update email to a client in a [matter type] case. Recent developments: [anonymized facts]. Use plain English, the firm's professional voice, and an honest tone about timeline. Do not promise outcomes. Mark anything that requires attorney verification in brackets. The attorney will review and send.
Honest about the line

Privilege, competence, and AI in legal practice

Ethics rules are not an afterthought on this page. ABA Model Rule 1.1 (competence) and recent state bar guidance now require attorneys to understand the AI tools they use in practice, including their limits. Rule 1.6 (confidentiality) means privileged client information cannot leak into a public AI tool that may train on or retain it. Mata v. Avianca made the rule about hallucinated citations crystal clear: courts sanction lawyers who cite AI-generated cases that do not exist. The standard for AI use in a law practice is not whether it is fast. It is whether the attorney remained competent, the client information stayed confidential, and the final work product met the duty of candor to the tribunal.

Privileged information stays out of public AI tools

Client information and privileged communications never go into ChatGPT, Claude, or any public AI tool. Use firm-deployed, confidentiality-protecting environments only. If you would not paste it into a public forum, do not paste it into a public AI.

Every AI-generated citation gets verified

Every case, statute, and regulation cited by an AI tool must be verified by an attorney in Westlaw or Lexis before filing. Mata v. Avianca established that courts sanction lawyers for hallucinated cases. AI citations are research leads, never authority.

Attorney review is required on every output

AI does not practice law. Every AI-drafted client communication, contract, or filing gets attorney review and sign-off before it leaves the firm. The output is junior-associate-level draft work, not final work product.

Disclose AI use per your jurisdiction

Disclose AI use to clients per your state bar's current guidance. Transparency is increasingly the rule, not the exception. Check your jurisdiction's most recent AI ethics opinion before settling on a default disclosure posture.

How attorneys use AI Monday morning

Six concrete uses you can put to work this week, with attorney review built into every step.

Attorney reviewing intake notes on a tablet at a desk

Client intake forms triaged before your first call

AI sorts the morning's new inquiries by practice area, surfaces conflict-check names, and pulls a five-bullet fact summary per matter. You walk into the first call already oriented.

Marked-up contract with notes in the margins

Contract redlines that flag clauses for your review

Run a first-pass review across an inbound contract. AI flags assignment, indemnity, term, and any unusual provisions. You review the flagged sections, not the entire document.

Open law book and laptop on a desk in a small firm library

Research memos that point to source citations

Ask a research question, get a structured memo with citations as leads. Pull every citation in Westlaw or Lexis before relying on it. AI narrows where you look. The verification stays with the attorney.

Letterhead and pen on a desk with a small firm office in the background

Demand letter drafts that need editing, not rewriting

Hand AI the facts and the target outline. Get a structured first draft with legal conclusions marked for verification. Your work becomes the editing and the strategy, not starting from a blank page.

Attorney typing a client email on a laptop

Status updates to clients in minutes, not hours

Draft routine client status emails in your firm voice. Plain English unless the client is sophisticated. Honest about timeline. Attorney reviews and sends. The afternoon comms backlog disappears.

Laptop with a firm website draft open on the screen

Marketing copy that passes state bar review on the first read

Firm website copy, blog posts, and email newsletters drafted with your jurisdiction's advertising rules built into the prompt. Attorney reviews, edits where needed, approves. The compliance review stops being a bottleneck.

Copy client update prompt

Try it yourself, draft a client status update

This produces a DRAFT for attorney review, not a finished message. Anonymize the facts before you paste them. Never put real client names, matter numbers, or privileged communications into a public AI tool. Use firm-deployed, confidentiality-protecting environments for actual client work.

Fill in your details

Practice area and matter category. No client names, no matter numbers.

Where the matter sits right now.

Anonymize all facts. No client names, opposing party names, or privileged communications.

What the firm will do over the next few weeks.

Used for context in the body if the firm includes hours in updates.

How your firm sounds in client emails. Two or three sentences.

Calibrates how much plain-English explanation the message includes.

Your prompt

live preview
You are drafting a status update email to a client of a small law firm. The output is a DRAFT for attorney review, not a finished message. The attorney will edit and send.

Matter context:
- Matter type (anonymized): {Commercial litigation, breach of contract claim}
- Case phase: {Discovery, pre-trial motions, settlement negotiation}
- Recent developments (anonymized, no privileged detail): {Opposing counsel produced documents on schedule. Two depositions completed. Settlement offer received and under review.}
- Next steps in the matter: {Respond to settlement offer within 10 days, complete remaining depositions, prepare summary judgment motion}
- Billable hours this period: {12.4 hours since last update}
- Firm voice and tone: {Direct, plain English, professional but warm. We avoid legalese and lecture tone.}
- Client sophistication level: {Business owner familiar with litigation, individual not familiar with the legal process}

Write a status update email with this structure:
1. Greeting that fits the firm voice.
2. A clear status summary, two to four sentences, that includes the recent developments in two to three sentences.
3. The next steps in the matter, in plain language, with realistic timelines.
4. Any input or decision needed from the client, asked for directly.
5. A professional sign-off in the firm voice.

Rules:
- Match the firm voice. Use plain English unless the client is sophisticated.
- Be honest about timelines and likely outcomes. Do not promise outcomes.
- Include the recent developments in two to three sentences.
- Mark anything that requires attorney verification in brackets like [VERIFY].
- Do not include legal advice, conclusions, or strategy that should come from the attorney directly.
- Output: greeting, status summary, next steps, ask for client input where needed, sign-off.
- Label the entire output DRAFT - FOR ATTORNEY REVIEW at the top.
Open in Claude

Frequently asked

Can attorneys use AI without violating ABA Rules 1.1 or 1.6?

Yes, when used correctly. ABA Model Rule 1.1 (competence) now requires attorneys to understand the benefits and risks of relevant technology, and recent state bar guidance treats AI literacy as part of that duty. Rule 1.6 (confidentiality) means privileged client information never gets pasted into a public AI tool like ChatGPT or Claude.ai on a consumer plan. For anything touching client matters, use confidential, firm-deployed environments only. The duty does not change. The tool does.

Will AI replace lawyers, paralegals, and associates?

No, not for the judgment work or the client relationship. A client deciding whether to trust your firm still reads the attorney across the table. AI is good at the load that keeps lawyers from doing that work: intake triage, document review first-pass, citation lookup (verified), drafting templates, and routine correspondence. The firms winning with AI are using it to free up attorney time for counsel and strategy, not replace it.

Is it ethical to use AI for legal drafting and research?

Yes, with conditions. Every citation gets verified against the actual case before it goes in a brief. Mata v. Avianca (S.D.N.Y. 2023) is the canonical warning here: Judge Castel sanctioned the attorneys $5,000 for filing a brief with six AI-hallucinated cases. The attorney signs and reviews every output, and clients are informed when AI materially helped, per the jurisdiction's guidance. AI drafts. The attorney practices law.

What AI tool should a small firm start with?

Legal-specific tools designed for attorney workflows and confidentiality protection. Lexis+ AI, Westlaw Precision AI, Harvey, and Thomson Reuters CoCounsel are built for legal research, drafting, and document review with appropriate data protections. Never paste client information, privileged communications, or matter details into public ChatGPT or Claude.ai on a consumer plan. For non-privileged work like marketing copy or internal templates, consumer tools are fine. The tool matters less than the rule about what goes into it.

How long does it take to learn AI for a law practice?

About 30 minutes to start using it for non-privileged tasks: general legal research questions, marketing copy, internal templates. About 2 to 3 weeks to set up a confidentiality-protecting workflow inside a firm-deployed environment for document review, drafting, and intake. ROI usually hits within the first month, most often for firms doing intake-heavy or research-heavy work where reclaimed attorney hours show up immediately on the realization rate.

Should I tell clients I use AI?

Yes, increasingly. Many state bars and the ABA are moving toward disclosure expectations, and Rule 1.4 (communication) already requires keeping clients reasonably informed about how their matter is being handled. Disclosure builds trust and matches where the profession is headed. The disclosure does not have to be heavy. A short line in the engagement letter that AI assists with research and drafting under attorney review is enough for most matters.

Can a small firm hire you to build something custom?

Yes. We build confidentiality-protecting AI workflows for solo and small firms with practice-management integration where it makes sense, Clio, Smokeball, Practice Panther, MyCase. Common builds include intake triage and conflict-check tools, document review first-pass, drafting libraries in your firm's voice, and verified citation pipelines. Free 30-minute scoping call to see if we are a fit. The contact form below routes the inquiry directly.

Want one built for your firm?

We build custom AI workflows for solo and small firm practices, tied to the tools you already use: Clio, Smokeball, Practice Panther, MyCase, Westlaw, and Lexis. Deployments are confidentiality-protecting by design, with attorney review built into every output. Free 30-minute scoping call to walk through your intake, doc review, drafting, or client comms bottleneck.