BlogAI use cases
AI use cases for law firms in 2026
The use cases that pay first are in the back office: Caddi opens matters from intake email, drafts time entries from email and calls, and files documents to iManage or NetDocuments.
The three worth automating this year, and why the back office comes before the drafting copilot.
- 99%of agent runs complete successfully
- ~45 minon average to build your first agent
- 1M+back-office tasks completed
By Jason AlafganiUpdated Oct 15, 2026
01Intake
Every new matter starts with someone retyping an email.
Referrals, web forms and demand letters arrive in a dozen formats. Someone reads each one and keys the client, the matter type, the other side and the deadline into the practice management system.
intake@ · ReferralNew client: lease dispute, letter attachedDemand letter.pdf
New matter
Matter opened in Clio, nothing retyped
A paralegal copies fields from the inbox into Clio, one inquiry at a time.
Caddi reads each inquiry and its attachments, pulls the parties, matter type and key dates, and opens the matter in Clio.
02Time and billing
The time that never gets billed is the cheapest revenue a firm has.
Short emails and quick calls fade by the end of the day. Rebuilt at month-end, they shrink or disappear, and no report ever shows them missing.
Draft time entries · Tuesday
- Email to opposing counsel re: extension1042 Brightwater0.2
Call with client re: lease terms1042 Brightwater0.4- Review court noticeMatter unsureReview
Drafted into Clio for the attorney to confirm
Reconstruct the month from memory, then chase the stragglers before the bill run.
Caddi drafts entries from the email and calls already in Outlook and RingCentral, matches them to the matter, and flags anything unsure for the attorney to confirm.
03Documents
Documents move between systems all day, by hand.
Read the PDF, rename it to the firm's convention, file it to the right workspace, and copy the key fields into the case record. The work never fits inside one tool, so it stays manual.
Read
Lease amendmentSigned Sep 30Brightwater
1042 Brightwater / Correspondence2026-09-30 Lease amendment (signed).pdfDownload, rename, upload, retype, repeated for every document that comes in.
Caddi reads each document, names and files it to the matter in iManage or NetDocuments, and fills the fields in the case record. Every run is logged.
From law firm teams running Caddi
“I used the Caddi time entries for submitting my time at the end of the month, and they were really good. I ended up saving a ton of time.”
Partner, law firm “It's able to look at an email and, even if it doesn't have the client matter number in it, figure out how that should be inserted into my timesheets. I really like that.”
Partner, law firm “With the last two cases that I tested, results were both good. In terms of the extraction, everything is good.”
Case assistant, law firm
Caddi
See how Caddi AI Agents can open matters and file documents and run your firm's back office the same way every time
Frequently asked questions
What are the highest-ROI AI use cases for law firms in 2026?
The fastest payback comes from the document- and inbox-heavy back office: client and matter intake, conflicts checks and onboarding, time capture and billing, collections follow-ups, document filing across the DMS, and mailbox and docket triage. Practice-side use cases (drafting, legal research, contract review) are higher status but slower to govern. Clio's 2026 data finds up to 74% of hourly-billed tasks are automatable with tools available today, and back-office billing capture alone can recover $65,000 to $162,000 a year for a five-attorney firm.
What are the main AI use cases for law firms?
Seven come up most. 1. Client and matter intake: turning inquiries, web forms and referral PDFs into open matters without re-keying. 2. Conflicts checks and client onboarding: assembling the file, searching existing parties and flagging only the exceptions. 3. Time capture, billing and collections: drafting entries as work happens, reconciling pre-bills, posting payments and chasing overdue invoices. 4. Document drafting and review: drafting copilots and clause comparison. 5. Legal research and matter summaries: research, deposition and document summaries, first-draft memos. 6. Document movement across the DMS: extracting, renaming, filing and routing documents across iManage, NetDocuments, Clio, Filevine or SharePoint. 7. Mailbox triage and docketing: classifying and routing shared-inbox mail and keeping dockets and calendars current.
How much unbilled time does a law firm lose?
Clio's 2026 Legal Trends data puts the unbilled-time gap at 2 to 5 hours per attorney per week. Recovering even half of it is a $65,000 to $162,000 annual revenue increase for a five-attorney firm. Most of it is short work, emails and calls, that is forgotten by the time the timesheet is rebuilt.
How much time and revenue does AI actually save law firms?
Wolters Kluwer's 2026 Future Ready Lawyer report found over 90% of legal professionals now use at least one AI tool daily, and 62% save 6 to 20% of their weekly time. On revenue, Clio's 2026 Legal Trends data shows AI-using firms nearly doubled revenue over four years with only a ~50% increase in clients and matters: capacity gains, not headcount. The differential is widening: mid-sized firms using AI extensively are about 65% more productive on capacity than small firms.
Should a firm automate the back office or buy a drafting copilot first?
Sequence by where ROI lands first. Back-office operations have bounded inputs, lighter governance, and countable ROI (cycle time, non-billable hours), so they return a defensible number in a single quarter. Practice AI carries privilege and malpractice exposure and longer committee review. Most firms get faster, safer wins by automating intake, billing, and document movement first, then expanding into substantive work.
Where does practice AI like drafting and research fit?
It is real and widely used: more than half of in-house teams use AI contract review, corporate legal departments report contract cycle-time cuts of up to 40%, and research and summaries are among the most-used tools in the 2026 Wolters Kluwer survey. It is also where governance is hardest, because privilege and malpractice exposure sit in the work product. Firms that have already built trust in AI on back-office operations tend to adopt it faster.
Why does the AI advantage compound for law firms?
Each automated workflow frees capacity that funds the next one, and the foundation you build for intake makes billing and document movement cheaper to automate the year after. Firms with a two-year head start are measurably better at their AI-assisted workflows than a firm starting today. As Clio puts it, the productivity gap 'does not close on its own. It compounds.' Waiting concedes margin every quarter you stay manual.
How does Caddi automate law-firm workflows without a developer?
Caddi uses a record-to-code approach: a paralegal or ops admin screen-shares a workflow exactly as they do it today, Caddi writes it as deterministic code, and it runs unattended on a schedule with audit trails. It connects across the tools a firm already runs (iManage, NetDocuments, Clio, Filevine, Salesforce, Microsoft 365, DocuSign), so there's no platform migration and no bots to babysit.