For lawyers, in-house counsel and paralegals

AI course for lawyers who can't afford a confident wrong answer

You spend hours on first drafts, contract markups and client emails that follow the same pattern, and you have probably seen AI invent a case citation. This live practicum teaches you to use AI on that routine work with verification built in. In two weeks you leave with configured tools, a prompt library for your own matters and a clear rule for what never goes into a chatbot.

Enroll for $250 Next cohort: November 16–25 · 4 live 90-minute Zoom sessions

Where AI helps in everyday legal work

First-pass contract review

Ask AI to go through a contract with the checklist you already use: unusual clauses, missing definitions, one-sided indemnities, inconsistent dates. It speeds up the first read, but it doesn't know your client's priorities or the governing law, so every flag is a lead to check, not a conclusion.

Client letters in plain language

Turn a dense internal memo into a clear explanation of risks and options for a client who isn't a lawyer. You set the tone and the level of detail, and AI handles the restructuring. Read the result for anything that sounds more certain than your actual advice.

Research starting points

Use AI to map the issues, suggest search terms and outline arguments on both sides before you open your research database. Never cite a case, statute or article AI gives you until you have found and read it in a primary source, because fabricated citations look exactly like real ones.

Summaries and chronologies

Feed in a long correspondence thread, a transcript or a file of documents and get a dated chronology with references to the source page. Spot-check entries against the originals: models sometimes merge two events or shift a date.

Templates, checklists and internal know-how

Draft engagement letters, intake questionnaires, closing checklists and notes for junior colleagues from your own past work. Frequent, lower-risk work where AI saves real effort, provided the input is anonymized.

What you do in each session as a lawyer

1
Tools and the core skill
You compare how ChatGPT, Claude and Gemini handle long documents and precise wording, and choose a setup that fits your budget and your firm's rules. Then we rewrite a vague legal prompt into one with context, task, constraints and a sample, on your own tasks.
2
Writing, correspondence, personal brand
You build templates for client updates, demand letters, polite refusals and difficult conversations about fees or bad news. We also work on tone for different cultures and on removing the AI smell.
3
Documents, data, visuals
The core session for lawyers: reviewing a contract and preparing questions for the other side, catching numbers AI gets wrong, and researching a topic quickly. We also cover exactly what must never be uploaded: personal data, material under NDA, trade secrets and confidential client information.
4
Assistants, automation, assembling the system
You build your own assistant for one recurring task, such as a first-pass NDA review against your standard positions, with persistent instructions. Then you audit your week, map your personal AI system and plan the next 90 days.

Full course program →

Worked example: first-pass review of an NDA

A client sends a counterparty's mutual NDA and wants to know by tomorrow whether it is safe to sign. You want a structured first read before you go through it line by line.

You are an experienced commercial lawyer reviewing a mutual NDA for a client.
Context: my client is [the disclosing / receiving party], a [type of business]. The purpose of disclosure is [purpose]. The contract is governed by [governing law].
Task: review the text below and list the clauses that deviate from a balanced mutual NDA.
For each issue give: clause number, an exact quote, why it may be a problem for my client, and a suggested alternative wording.
Constraints: do not state what the law requires. Where the answer depends on the governing law, write "check against governing law".
If something is missing (term, return of information, exclusions), list it separately.
Format: a table, then three questions to send to the other side.
Anonymized text of the NDA:
[paste the text with names and identifying details removed]

You get a table of flagged clauses, missing provisions and draft questions for the counterparty. Check each quote against the text, since models sometimes paraphrase while claiming to quote, and test every suggested wording against the governing law. If the review feels generic, add your own standard NDA positions as examples and run it again.

What you leave the practicum with

A tool setup that matches your budget and your firm's confidentiality rules.
A personal prompt library for contract review, client letters, chronologies and research outlines, built on your own matters.
A verification routine for citations, quotes, dates and figures that you apply before anything leaves your desk.
One working assistant for a recurring legal task, with instructions you can refine over time.
A map of your AI system and a 90-day roadmap, plus the recordings, workbook and answer-checking checklist.

Where AI doesn't replace a lawyer

AI doesn't know your client, the deal history or what the other side will accept, and it can't take responsibility for advice. It produces plausible citations that don't exist and states legal positions with the same confidence whether they are right or wrong. Professional judgment, confidentiality and the final signature stay with you. The practicum teaches you to delegate drafting and sorting while keeping the reasoning and the checking in your hands.

FAQ

Is it safe to put client documents into an AI tool?

Not by default. We cover which settings reduce the risk, how to anonymize a document, and what must never be uploaded. Your firm's policy and professional conduct rules come first.

What about AI inventing case law?

The practicum treats it as a given, not a rare glitch. You learn to use AI for structure and drafting, not as a source of authority, and to verify every citation in a primary source.

Is the content tied to one legal system?

No. The skills are jurisdiction-neutral: setting the task, reviewing, checking. Whenever a result touches substantive law, the rule is the same: check it against the law that governs your contract or dispute.

I'm not technical. Will I keep up?

Yes. No coding is required, and the no-code block in the last session is advanced and optional.

What if it isn't right for my practice?

Attend the first session. If it's not a fit, you get a full refund. Assignments between sessions take 15 to 20 minutes on a real work task, so you find out quickly.

Ready-made kits for your role

If you need the tools today rather than on the cohort date: a prompt library and a self-paced course on the same program.

$189

Legal Document Review Prompt Kit

Contract review, clause comparison, risk summaries and client explanations — 120 prompts with guardrails.

What's inside →

$290

Documents, Data & Contracts with AI

Contract review, spreadsheets and numbers, research and presentations — where AI is reliable and where it lies. Self-paced text lessons with exercises and templates.