Use cases by industry6 use cases
AI automation for law firms that cannot afford a leak
A 25-lawyer litigation firm in Germany receives court mail through beA every day. A paralegal reads each document, works out the date of service, calculates the deadline with the right state's public holidays, and enters it in RA-MICRO or Advoware for a lawyer to confirm. A missed deadline is a malpractice claim, so the work is careful, repetitive and exactly what court deadline docketing takes over, with rules in code computing every date and a lawyer confirming each one.
The money in legal AI has gone to US plaintiff firms: EvenUp raised $150M at a valuation above $2B and Eve raised $103M at $1B, both for personal injury work, where medical chronologies and demand letters are now bought rather than built. Commercial and employment firms in Europe, and in-house teams, have different problems: intake with a conflict check before anyone calls the prospect back, and contract review against the firm's own playbook rather than a vendor's.
Confidentiality shapes every design. KPMG and the University of Melbourne found that 48% of employees had uploaded sensitive company information to public AI tools, and in a law firm that means client files. A private AI gateway with single sign-on, redaction and EU-hosted models is often the first thing I would build, because it removes the reason to use personal ChatGPT accounts.
What I would build for law firms
A company assistant that answers from SharePoint and Drive, and respects who may see what
Answers staff questions in Teams and Slack from SharePoint, Drive and Confluence, cites every source, and only searches what the person asking may open.
SharePoint and Teams / Google Drive / Slack / Confluence or Notion / Microsoft Entra ID or Google groups
Law firm intake that runs the conflict check before anyone calls the prospect back
Captures every inquiry, after hours too, searches conflicts across parties and their groups, flags deadlines, and prepares the engagement letter for a lawyer.
Practice management system (RA-MICRO, Clio, LEAP or Actionstep) / Outlook and Microsoft 365 / Website form and chat / Calendar / iManage
A private ChatGPT for every employee, behind single sign-on, with redaction and team budgets
A company chat app behind single sign-on that strips client data and secrets, routes to approved EU or self-hosted models, logs use, and gives each team a budget.
Microsoft Entra ID or Google Workspace / Slack or Teams / OpenAI, Anthropic and Mistral / Open models on EU infrastructure / Open WebUI, LibreChat or a custom interface
Commercial leases turned into key dates and clauses, with a lawyer checking the ones that matter
Turns a portfolio of commercial leases and their amendments into cited abstracts and computed deadlines in Yardi or MRI, with a lawyer confirming breaks and options.
SharePoint / Yardi Voyager or MRI / Excel / Critical-dates calendar / Index publishers (Destatis, CBS, ONS)
Contract review against your own playbook, with every flag tied to the clause that caused it
Checks each clause of a customer's DPA or NDA against your playbook, quotes the text behind every flag, and drafts tracked changes that a lawyer accepts or rejects.
Microsoft Word and SharePoint / iManage or NetDocuments / Clio / Ironclad or Juro, where a CLM exists / The playbook and clause library
Court mail read, deadlines calculated by rules, and every one confirmed by a lawyer
Reads incoming court documents, calculates each deadline with tested rules rather than a model, proposes entries in your case system and waits for a lawyer to confirm.
beA or e-filing notices / RA-MICRO, Advoware or DATEV Anwalt / Clio, LEAP or Actionstep / Outlook / Holiday and rules calendar
The legal work that is really data handling
Court mail into the case file
Judgments, orders and hearing notices arrive through beA in Germany or as e-filing notices elsewhere. Classifying each one, extracting the date of service and matching it to the right matter is extraction work; the deadline arithmetic stays in code.
Conflicts hidden in group structures
A prospect's opponent may be a subsidiary of an existing client. Fuzzy matching against parties and related entities finds the candidates, and a person confirms every possible conflict before the firm engages.
E-billing rejections
Invoices to corporate clients bounce when time narratives or LEDES and UTBMS codes break the client's billing guidelines. Checking entries against each client's rules before submission avoids the write-downs.
Subject access requests
GDPR access requests, frequent in employment disputes, mean finding, reviewing and redacting large volumes of email. Search and first-pass redaction can be automated; a lawyer decides what is released.
Professional duties the build is designed around
A lawyer confirms every deadline
The system proposes deadlines and alerts the responsible lawyer; it never enters one as final. The four-eyes principle survives the automation, and every calculation shows its inputs.
No legal advice from an intake bot
An intake assistant collects facts, flags limitation risks and books the consultation. It says it is an AI, as Article 50 of the EU AI Act has required since 2 August 2026, and it never tells a prospect whether they have a case.
Client data stays where it is allowed
Documents are processed on EU infrastructure or self-hosted open models, under a processing agreement, with retention you set. Matter-level permissions from iManage or NetDocuments are enforced at query time.
Buy the review tools when they fit
Harvey, Spellbook, Luminance and LegalOn review contracts well. Custom work wins when the value is your own playbook and precedents, your DMS, and control over where the data is processed.
Software these builds usually connect to
The systems do not get replaced. The build sits across them, reads from them through their APIs, and writes results back.
Clio / iManage / NetDocuments / beA / RA-MICRO / Advoware / DATEV Anwalt / Actaport / LEAP / Actionstep / Litify / Filevine
Frequently asked questions
Can AI calculate court deadlines for a law firm?
It can read the court document, find the date of service and propose the deadline, but the arithmetic should be rules in code: procedural periods, weekends and public holidays by state or region. The model reads and matches, the rules compute, and a lawyer confirms every entry before it becomes final. That keeps the four-eyes principle and makes each deadline explainable.
Is it safe for a law firm to use ChatGPT with client documents?
Not through personal accounts. The safer setup is a firm gateway behind single sign-on that routes to approved models under business terms, EU endpoints or self-hosted open models, redacts personal data before anything leaves, and logs use. Consumer tools used without that layer are how client files end up outside the firm's control, whatever the policy says.
What can an AI intake assistant do for a law firm?
Turn web forms, emails and after-hours calls into structured intake: matter type, parties, jurisdiction and urgent limitation dates. It runs the conflict search, triages against your practice-area criteria, drafts the engagement letter and books the consultation. It discloses that it is an AI and never gives legal advice, and a lawyer accepts every matter.
Harvey or a custom build: which fits a mid-sized firm?
If you want a broad legal assistant and your data may go where the vendor processes it, a product like Harvey or Luminance is the faster route. A custom build fits when the value is your own playbook and precedents, when documents must stay in your DMS or on EU or self-hosted infrastructure, or when the work is firm-specific plumbing such as docketing and conflicts.