Home · Professional services · AI automations
AI for small law firms: 8 real automations from solo lawyers and law firms
Updated October 18, 2026 · 8 sourced examples · no paid placement
Short answer: Solo lawyers and small firms here mostly use AI on the work around the legal work: a second look at intake decisions, case notes from recorded client conversations, replies to opposing counsel, follow-up emails after networking events, a daily summary of the inbox and meeting action items, a weekly review of the firm's numbers, and blog and social posts. The AI is ChatGPT, Claude, Microsoft Copilot and Perplexity, not special legal software, though one firm also uses Lexis+ AI. Reported numbers are few: Todd McMurtry says he saves an hour or two every day, and Angel Reyes & Associates says editing a blog post now takes 20 to 30 minutes instead of about four hours. Every lawyer here has a person check the output, and two of them warn about confidentiality. These are what each firm reported, not promises.
Below are 8 documented examples from law firms, from a solo business lawyer in Austin to a seven-state personal-injury firm, told by the lawyers on podcasts and in a bar president's column, or reported in ABA Law Practice Magazine: who did it, what the AI does, what they reported and where the story came from. After the list: the jobs grouped by area, what to keep human, what they say about client confidentiality and how to try one this week. This is not legal or ethics advice; check your own state bar's guidance before client information goes into any AI tool.
Other industries too: see real AI automations by task for any small business.
What these 8 businesses actually did
- A Claude project that rechecks every intake decision: McCready Law (personal injury, seven states) (owner-told, via Maximum Lawyer podcast). A Claude project with the firm's screening criteria reviews each intake decision and rates the call. Internal AI bots answer SOP and handbook questions, and agents check case files at set points. Reported: Self-reported: a wrongly declined lead gets a call back. He suggests this intake review as the easiest first AI project for a small firm. Humans review everything, with an audit trail. Full case
- Recorded Spanish client conversations turned into case notes: Jaskot Law (press, via ABA Law Practice Magazine roundtable: All Things Leadership and AI (Mar/Apr 2025), reproduced by Lawyers Concerned for Lawyers Massachusetts). He records the conversations, has them transcribed and translated, then feeds the transcript to AI with an instruction that turns it into a case note for his practice-management system. The roundtable doesn't name the AI or transcription tools. Reported: "I have always struggled with taking good notes and now my notes are excellent thanks to transcription, translation and summarization," he says, adding that it saves time and gives more depth and detail than he thought he could get. No figures are given. Self-reported. Full case
- Replies to opposing counsel drafted in Claude, then made firmer or softer: Todd McMurtry (Fort Mitchell, Kentucky) (owner-told, via Kentucky Bar Association Bench & Bar, Sep/Oct 2025: President's Page, The Artificial Intelligence Revolution (Todd V. McMurtry)). He drops the letter into Claude, tells it which side he represents, how he wants to respond, the key facts and other relevant information, and asks for a responsive letter. Then he asks it to make the letter more firm or less aggressive. His firm uses Lexis+ AI plus Claude or ChatGPT as its main AI tools. Reported: He says redrafting the tone "takes seconds" and that he saves "an hour or two every day" with these tools. He warns that free versions aren't confidential, so client information goes only into a paid version, and that all AI work has to be checked carefully. Self-reported. Full case
- A business-card photo turned into a contact and a follow-up email: Zoll Firm (Austin, Texas) (press, via ABA Law Practice Magazine: Making Practical Use of AI in My Practice Today (Daniel E. Pinnington and Reid F. Trautz, Q&A with Mitch Zoll)). He photographs a business card with the ChatGPT phone app and his own custom GPT reads the card so he can pull it into his contacts, looks at the person's role for a business connection to his practice, and drafts a personalized follow-up email for him to send. It is told never to open with "I hope this finds you well" and to offer to connect without being pushy about his services. Reported: He describes it as one of the agents he uses daily and says the emails read "just like an email I would write after a networking event." He reviews and sends each one himself. No figures are given. Self-reported. Full case
- A daily inbox summary and meeting action items in Copilot: Zoll Firm (press, via ABA Law Practice Magazine: Making Practical Use of AI in My Practice Today (Daniel E. Pinnington and Reid F. Trautz, Q&A with Mitch Zoll)). He uses a Copilot agent that stays inside his Microsoft environment to protect client information: it scans his Outlook email, summarizes it grouped by sender and subject and lists open action items. In Teams, Copilot transcribes his meetings and gives a summary and action items, which go into his client call logs. He also has Copilot check meeting times people offer against his calendar and draft the reply. Reported: He says the daily summary helps him organize the day, set the next day's priorities and keep up with billable time, and the call summaries are "invaluable when a client calls back a month later." No figures are given. Self-reported. Full case
- A Monday review with AI "CMO, CEO and CFO" agents, and Claude on QuickBooks: Business Law Group (New Orleans) (owner-told, via Endeavors AI podcast: Above the AI Cloud, how a boutique New Orleans firm went AI-first (Amanda Butler Schley)). Every Monday she consults three AI 'agents' set up as a CMO, CEO and CFO, giving them the firm's real reports. She uploads transcripts of meetings with her human outsourced CFO so the AI's advice stays consistent with his. Claude is connected to QuickBooks for financial analysis. An AI receptionist is in testing. For content she now questions every sentence, after another lawyer flagged the invented law in a published post and she had to pull it. Reported: Self-reported, no numbers: she says connecting Claude to QuickBooks unlocked analysis the firm never had. Lessons she shares: AI can 'say the law wrong' even on simple summaries; models tend to adopt your stance and feed it back (contrarian prompts haven't fixed that for her); the same question asked ten times gives different answers. Her team is skeptical of the AI receptionist so far. Full case
- Blog posts built from real intake questions, with a human editor on every post: Angel Reyes & Associates (owner-told, via Traffic to Leads podcast (JuvoLeads): Graham Griffin on AI content systems at Angel Reyes & Associates). The firm moved marketing in-house and plans topic clusters (one insurance cluster is 150 articles) around the questions people actually ask at intake, plus Google 'People Also Ask' data and SEO research. A six-step pipeline of custom GPTs drafts a full post in about three minutes, with Claude used for language and Gemini for visuals through their APIs, and an 8,000-word style prompt that holds the firm's voice. The last step is always a human editor. Strong posts are repurposed into email campaigns and social posts. Reported: Human time per post fell from about 4 hours to 20–30 minutes of editing; testing roughly 100 posts cost about $15 in API fees; organic users rose 2–3x in about a year. His bio also credits his work with contributing to a 120% increase in web-driven cases over eight months (self-reported). The pipeline itself isn't published. Full case
- Plain-language blog and social posts, with Claude for writing and Perplexity for research: Berardi Immigration Law (press, via ABA Law Practice Magazine roundtable: All Things Leadership and AI (Mar/Apr 2025), reproduced by Lawyers Concerned for Lawyers Massachusetts). After evaluating several platforms, the firm settled on Claude for legal writing and analysis and Perplexity for research with current, cited sources. It uses AI to analyze trends, spot emerging issues and write first drafts of blog and social content that turns complex legal concepts into accessible posts; lawyers verify and customize every draft. Reported: She calls content creation one of her favorite uses and says AI has helped spot analytical blind spots. All AI output goes through what she calls a strict verification protocol, and client information is redacted before it enters any AI system. No figures are given. Self-reported. Full case
These are each company's own results with their own tools and teams, as reported by the source. Vendor case studies are written by the company selling the tool. None of them are promises.
What do small law firms automate with AI?
Intake and new matters. At McCready Law, a Claude project loaded with the firm's screening criteria reviews each intake decision and rates the call, so a wrongly declined lead gets a call back; Michael McCready suggests this as the easiest first AI project for a small firm. Jared Jaskot of Jaskot Law holds most client conversations in Spanish; he records them, has them transcribed and translated, and has AI turn the transcript into a case note for his practice-management system. He says his notes went from a struggle to "excellent".
Letters and email. Todd McMurtry drops a long letter from opposing counsel into Claude, says which side he represents, how he wants to respond and the key facts, asks for a reply, then asks for a firmer or softer version. After networking events, Mitch Zoll of Zoll Firm photographs each business card in the ChatGPT app and gets the contact details plus a personal follow-up email drafted for him to edit and send.
The inbox, meetings and the calendar. Zoll has a Copilot agent summarize his Outlook email each day by sender and subject with open action items, uses Copilot in Teams for meeting summaries and action items that go into his client call logs, and has Copilot check offered meeting times against his calendar. He keeps this inside Microsoft 365 so client email stays in his own Microsoft environment.
Running the firm. Every Monday, Amanda Butler Schley of Business Law Group consults three AI "agents" set up as a CMO, CEO and CFO and gives them the firm's real reports, and she says connecting Claude to QuickBooks unlocked financial analysis the firm never had. At McCready Law, internal AI bots answer staff questions about SOPs and the handbook, and agents check case files at set points.
Marketing. Angel Reyes & Associates plans blog topic clusters around the questions people ask at intake; a six-step pipeline drafts each post with Claude for language and Gemini for visuals, and a human editor is always the last step. The firm reports human time per post fell from about four hours to 20 to 30 minutes and organic users rose 2 to 3 times in about a year. Rosanna Berardi's immigration firm uses Claude for legal writing and Perplexity for research, and uses AI to turn complex legal concepts into first drafts of blog and social posts that a lawyer verifies.
Phones, as one example among many. The law-firm phone and intake-agent examples on this site (Ghaffari Law Firm, STG Divorce Law and Grolman LLP) are vendor case studies, so they are on the law-firm receptionist guide, linked below, rather than in this list.
What should stay human?
The law itself. Schley published a blog post built on a ChatGPT summary of a CLE that invented the law, and pulled it after another lawyer flagged it; she now questions every sentence. Zoll's advice is "Use AI to keep up—not to catch up or give up," and he tells of a lawyer who called an AI-drafted response to a motion "ready to file" after 15 minutes: "That is not how to use AI."
Anything that goes out under your name. McMurtry says all AI work has to be checked carefully. Berardi's firm runs all AI output through what she calls a strict verification protocol, Angel Reyes keeps a human editor as the last step on every post, and McCready says humans review everything at his firm, with an audit trail.
What do these lawyers say about client confidentiality?
Free plans. McMurtry warns that free versions of these tools are not confidential, so he says not to share client information unless you use a paid version.
Where the data goes. Zoll points out that documents loaded into a custom GPT go onto OpenAI's servers, which is why his inbox summaries run in Copilot inside his Microsoft tenant. Berardi's firm uses only enterprise-grade AI tools and redacts client information before it enters any AI system. This page is not legal or ethics advice; the ABA (Formal Opinion 512) and several state bars have issued guidance on lawyers' use of AI.
Do you need special legal AI software?
Not for the jobs on this page. They run on ChatGPT, Claude, Microsoft Copilot and Perplexity. McMurtry's firm adds Lexis+ AI for legal research, and McCready Law keeps its Smart Advocate case-management system alongside Claude. Research and citations are where these firms lean on legal research tools and their own checking.
Tools named in these examples
| Tool | Starting price (base plan) | Checked |
|---|---|---|
| ChatGPT | Free $0; Go $8/mo; Plus $20/mo; Pro $100/mo | 2026-10-17 |
| Claude | $17/mo billed annually ($20 monthly) | 2026-10-18 |
| Microsoft Copilot | Individuals: Microsoft 365 Personal with Copilot $9.99/mo; Business: Copilot add-on from $18/user/mo paid yearly (promo, normally $21) | 2026-10-17 |
| Perplexity | Pro $17/mo billed annually (promo shown on Pro page) | 2026-10-18 |
Starting prices from each vendor's pricing page on the date shown. AI phone and texting features can be priced separately from the base plan, so check before you buy. Also named in these examples: Google Gemini and QuickBooks (prices on the tools page), plus Lexis+ AI, Smart Advocate, Outlook and Microsoft Teams (prices not checked here).
How to start this week
- Pick the one job that eats non-billable time: intake decisions, case notes, routine letters, follow-up emails, the morning inbox or blog posts.
- Use a paid plan before any client information goes in, as McMurtry advises, and check your state bar's guidance. If your firm runs on Microsoft 365, Copilot keeps the work inside your own Microsoft account, as Zoll does.
- Save a project in ChatGPT or Claude with what the job needs: your screening criteria (McCready Law), a case note you wrote yourself (Jaskot Law), past letters for tone (McMurtry) or your style guide (Angel Reyes & Associates).
- Have a lawyer read every output before it reaches a client, a court, opposing counsel or your website, and verify every legal point yourself.
- Once the drafts are reliably close, make it routine: every intake decision rechecked, every client meeting turned into a note, every networking contact followed up.
Have Claude review every intake decision so declined leads get a second look
A Claude project with the firm's screening criteria reviews each intake decision and rates the call.
Turn recorded Spanish client conversations into case notes for your practice-management system
He records the conversations, has them transcribed and translated, then feeds the transcript to AI with an instruction that turns it into…
Quoted: Jared Jaskot, owner, Jaskot Law · via ABA Law Practice Magazine roundtable: All Things Leadership and AI (Mar/Apr 2025), reproduced by Lawyers Concerned for Lawyers Massachusetts · 2025-03-01Draft the reply to a long letter from opposing counsel in Claude, then adjust the tone
He drops the letter into Claude, tells it which side he represents, how he wants to respond, the key facts and other relevant information…
Snap a business card after a networking event and get the contact plus a follow-up email drafted
He photographs a business card with the ChatGPT phone app and his own custom GPT reads the card so he can pull it into his contacts, look…
Get a daily summary of your inbox and meeting action items with Copilot, inside Microsoft 365
He uses a Copilot agent that stays inside his Microsoft environment to protect client information: it scans his Outlook email, summarizes…
Run a weekly review with an AI 'C-suite' (CMO, CEO, CFO) fed your real reports, with Claude connected to QuickBooks
Every Monday she consults three AI 'agents' set up as a CMO, CEO and CFO, giving them the firm's real reports.
Turn real intake questions into blog clusters with a multi-step custom GPT pipeline and a human editor on every post
The firm moved marketing in-house and plans topic clusters (one insurance cluster is 150 articles) around the questions people actually a…
Turn complex immigration-law topics into blog and social posts with Claude and Perplexity
After evaluating several platforms, the firm settled on Claude for legal writing and analysis and Perplexity for research with current, c…
Questions
How are small law firms using AI?
In these examples: a second look at intake decisions (McCready Law), case notes from recorded Spanish client conversations (Jaskot Law), replies to opposing counsel (Todd McMurtry), follow-up emails after networking events and a daily inbox and meeting summary (Zoll Firm), a weekly review of the firm's reports and QuickBooks analysis (Business Law Group), and blog and social posts (Angel Reyes & Associates, Berardi Immigration Law).
What is the best AI for a small law firm?
The lawyers here use general assistants: ChatGPT, Claude, Microsoft Copilot and Perplexity, and one firm adds Lexis+ AI for research. ChatGPT is listed from Free $0 (Plus $20/mo), Claude Pro at $17/mo billed annually ($20 monthly), Perplexity Pro at $17/mo billed annually, and the Microsoft Copilot business add-on from $18/user/mo paid yearly (promo, normally $21). Pick the one that fits the job and where your client data already lives.
Can lawyers put client information into ChatGPT or Claude?
The lawyers here are careful. Todd McMurtry says free versions are not confidential, so client information goes only into a paid version; Mitch Zoll keeps client email inside Microsoft Copilot; Berardi Immigration Law uses enterprise-grade tools and redacts client information first. This is not legal or ethics advice; check ABA Formal Opinion 512 and your state bar's guidance.
Can AI help with law firm intake?
Yes, behind the scenes. McCready Law has a Claude project recheck every intake decision so a wrongly declined lead gets a call back, and Jaskot Law turns recorded client conversations into case notes. AI intake agents on the phones appear only in vendor case studies (Ghaffari Law Firm, STG Divorce Law), which are on the law-firm receptionist guide.
How do law firms use AI for marketing?
Angel Reyes & Associates drafts blog posts around real intake questions with a multi-step pipeline and a human editor (about four hours per post down to 20 to 30 minutes of editing, self-reported), and Berardi Immigration Law uses Claude for writing and Perplexity for research, with AI drafting blog and social posts a lawyer verifies. Business Law Group's warning: a ChatGPT summary of a CLE invented the law in a published post.
How much time does AI save a law firm?
Few of these lawyers give numbers. Todd McMurtry says he saves an hour or two every day with Lexis+ AI plus Claude or ChatGPT, and Angel Reyes & Associates reports the blog-post figures above. Zoll, Jaskot, Schley, McCready and Berardi describe the benefit without figures.
Can someone set it up for us?
Yes. The $197 Quick-Win sets up one automation in accounts you own, live within 48 hours of getting access, with a refund if it is not live in 48 hours or is not doing its job in the first 14 days.
Related guides
- AI automations for any small business, by task
- AI automations for small accounting firms and bookkeepers
- AI automations for small insurance agencies
- AI automations for small marketing agencies and consultants
- AI client intake and onboarding automation: real examples
- AI email triage and inbox management: real examples
- AI receptionist for law firms, insurance agencies and advisors
- AI operations tools for law and accounting firms
- AI marketing tools for law firms and service businesses
- All professional services AI examples
- Have one automation set up for you ($197)
One short email a day: 3 new AI use cases from real small businesses, each linked to its source, plus a Tool of the Day. Free, no spam, unsubscribe anytime.
Want one set up for you?
The $197 Quick-Win: we set up one AI automation that saves you hours each week, in accounts you own, live within 48 hours of getting access. No call needed. Refund if it isn't live in 48 hours or isn't doing its job in the first 14 days.
See the $197 Quick-Win Bigger setups