Skip to content

Lesson 1 of 9 in AI in legal and professional services

Which professional work AI drafts, and how much review each piece gets

Route one piece of work: let teams self-serve the routine, start the expert at the AI draft, and set review depth by the stakes.

2-minute read, 20-minute exercise. Updated . Free lesson, no account needed.

You leave with: AI draft and review routing card (blank card, worked example), three AI prompts.Blank card (PDF)The card as text, with a worked example

In 60 seconds

  • Encode the routine, then hand it off. With contracts and positions encoded in a platform, other teams serve themselves instead of queuing for legal.
  • Start the expert at AI's draft. Summarizing, drafting and search are generative AI's three ways into professional work; no one starts from a blank page.
  • Match the review to the stakes. A routine contract redraft can be accepted after a scan of the redlines; a $100 million negotiation or a regulatory response warrants close review of every AI suggestion and the decisions that follow.
  • Verify before any handoff. With diagnostics run and each figure traced to its source, the professional verifies every fact before signing.

In its 2026 Legal Trends Report, Clio, a legal software vendor, reports that 85% of legal professionals have some level of AI adoption at their firms. With adoption that wide, what you still decide is the routing. Which work does AI draft, and how much review does each piece get?

A free lesson from AI in legal and professional services, one of its 9 lessons. Read another free lesson

The routing tree

Take the next piece of professional work that crosses your desk and walk it down these gates.

  1. Encoded and routine? Contracts and positions already live in a platform.
    • If yes: other teams self-serve. Sales can create contracts without waiting for legal. Procurement gains insight into purchase orders and renewals.
    • If no: step 2.
  2. A draft, a summary, a search?
    • If yes: the professional starts from the AI version.
  3. How high are the stakes?
    • HIGH: a $100 million negotiation, a regulatory response, then close review of every AI suggestion, and of the decisions that follow.
    • ROUTINE: a contract redraft, then scan the redlines; it can be accepted.
  4. Before any handoff: diagnostics run, proof of work shown, every fact verified by the professional who signs.

Gate one opens only after encoding: your contracts and positions captured in a platform. Once they are, much legal work can shift to other teams. Whether one agent should carry a whole engagement is a separate decision. This gate routes single, routine pieces.

Gate two holds the three ways generative AI helps professional services: summarizing, drafting, search. Summarizing cases and documents works when the right content is fed in. Drafting gives a strong first version, so no one starts from a blank page. Search explores, especially on models trained on the domain's text. Document review needs more depth than this tree gives. Audit work does too.

Gate three sets the depth of review by the stakes. A routine contract redraft can be accepted after a scan of the redlines. A $100 million negotiation or a regulatory response warrants close review of every AI suggestion, and of the decisions that follow. Sizing that review is its own lesson. This gate only decides that it happens.

Before any handoff

The fourth gate catches every route. Wherever a piece of work heads for a signature, the checks travel with it. AI can draft a convincing letter in any tone. The polish is the trap: the letter is a starting point, not ready to ship. Verifying every fact, and taking it over the finish line, belong to the professional who signs.

Before the work reaches the signer, have the system run dozens of diagnostic checks and year-over-year anomaly checks. Require proof of work for each figure: the exact box or number in the source document that fed it. Earlier generations of professional software did this tick-and-tie, and users expect it.

Exercise (20 minutes): route one piece of real work

Take the next piece of professional work that crosses your desk, or one sitting in your queue this week, and route it down the card's gates: encoded, AI-drafted, stakes, handoff checks. You won't know every line, and that is the point of filling it in. Every line you can't fill gets an owner and a date before you close the card. The card download includes a filled-in example of a contract renewal that is not ready to hand off yet, and why.

AI draft and review routing card (PDF)

A blank card and a filled-in example, one page each.

Open the card (PDF)
Show the three prompts: fill the card, test it, turn it into a plan

Use AI to fill in and test your card

These three prompts go into your company's approved AI assistant. Use the first to fill in the card one row at a time; it won't guess a number for you. Use the second when the card is done, to hear the case against handing the work off. The third asks about your industry and your piece of work and turns the lesson into a 30-day plan, with a checklist, rubric or brief for your boss if you want one. Use placeholders for company, system and people names; the help doesn't depend on them.

Prompt 1: Fill in my card with me

Paste this, then add your piece of work in one or two sentences and the card below.

Prompt
Before you start: use your company's approved AI assistant, not a personal account. Replace company, system, supplier and people names with placeholders (Company A, System X) and round any figures. The help you get does not depend on them.

You are helping me fill in a routing card for one piece of professional work. The card is pasted at the end.

Rules:
- Ask me about one row at a time, in order, and wait for my answer before moving on.
- Write down only what I tell you. Never suggest, estimate or invent a number, a name, a yes/no answer or a decision for me, even if I ask you to guess. If I ask, say that this row needs a real answer and ask who could give it.
- If I don't know an answer, write "Not known" and ask who should find out and by when. Put that in the Owner / by when column.
- After the last row, show the whole card as a table with three columns: Question, Answer, Owner / by when. Then list every row marked "Not known" with its owner, and remind me that any unanswered row means the work is not ready to hand off.

My piece of work in one or two sentences: [DESCRIBE THE WORK]

The card:
[PASTE THE CARD]

Prompt from The AI Decision Course by Dr. Tony Hoang, host of The Artificial Intelligence Podcast (Spotify, Apple Podcasts): tonyphoang.com/learn

Prompt 2: Stress-test my card

Paste this, then your filled card.

Prompt
Before you start: use your company's approved AI assistant, not a personal account. Replace company, system, supplier and people names with placeholders (Company A, System X) and round any figures. The review does not depend on them.

Below is a routing card I filled in for one piece of professional work. Act as a skeptical reviewer at the decision meeting and argue the case for HOLD, NOT READY.

For each row:
- Say whether the answer is specific enough to act on (a number, a named role, a date) or too vague.
- Point out where answers contradict each other, for example a routine rating against a review depth that reads every suggestion, or self-serve routing while the platform still queues work at legal.
- Say what evidence you would ask for before accepting it.

Do not rewrite my answers and do not invent facts about my work. If a row is blank or says "Not known", treat it as not ready.

Finish with: your recommendation (SELF-SERVE, AI DRAFT + REDLINE SCAN, AI DRAFT + CLOSE REVIEW, or HOLD, NOT READY), the three rows that most need work, and one question I should ask my team on Monday.

My card:
[PASTE YOUR FILLED CARD]

Prompt from The AI Decision Course by Dr. Tony Hoang, host of The Artificial Intelligence Podcast (Spotify, Apple Podcasts): tonyphoang.com/learn

To paste the card into prompt 1, copy its rows from the PDF.

Prompt 3: Make it yours

Paste this and answer its four questions.

Prompt
Before you start: use your company's approved AI assistant, not a personal account. Replace company, system, supplier and people names with placeholders (Company A, System X, Supplier B).

You are helping a leader apply one lesson to their own situation. The lesson's four decisions are below. Use only them and what I tell you. Never invent facts, numbers, vendors, laws or results about my organization or my industry. Where a number is needed and I have not given one, say what to measure and who should measure it, instead of guessing.

First, ask me these four questions, one at a time, and wait for each answer:
1. My industry, and roughly how many people work in my organization.
2. The piece of professional work I want to route, in one or two sentences.
3. Where it stands today: queued, being drafted, or about to go out for signature.
4. What worries me most about it.

Then give me the following. Keep it short and practical:
A. My plan. For each of the four decisions: one action written for my piece of work, the role that should own it (a role, not a name), and what should be done within 30 days. If my industry makes a decision harder, for example regulated data, client-facing work or professional sign-off, say which decision and why in one sentence.
B. Three questions to ask my team on Monday, specific to my piece of work.
C. Ask me which one I want: a checklist for routing our work, a rubric to score review depth, or a one-page brief for my boss. Then write it, and finish with a prompt I can paste later to update it.
Mark anything that rests on an assumption with "(check)".

The four decisions:
1. Encode the routine, then hand it off. With contracts and positions encoded in a platform, other teams serve themselves instead of queuing for legal.
2. Start the expert at AI's draft. Summarizing, drafting and search are generative AI's three ways into professional work; no one starts from a blank page.
3. Match the review to the stakes. A routine contract redraft can be accepted after a scan of the redlines; a $100 million negotiation or a regulatory response warrants close review of every AI suggestion and the decisions that follow.
4. Verify before any handoff. With diagnostics run and each figure traced to its source, the professional verifies every fact before signing.

Prompt from The AI Decision Course by Dr. Tony Hoang, host of The Artificial Intelligence Podcast (Spotify, Apple Podcasts): tonyphoang.com/learn

Your company's assistant may answer differently, so read what it writes before you use it.

Check yourself

Your sales lead wants to generate contracts from your standard positions. What has to exist first?

Show the answer

The contracts and positions encoded in a platform. Then sales can create contracts without waiting for legal.

An AI-drafted response to a regulator lands on your desk before a deadline. How closely do you read it?

Show the answer

Closely: high stakes warrant close review of every AI suggestion and the decisions that follow. A redline scan will not do.

A vendor demo shows figures with no link back to the source documents. What do you ask for?

Show the answer

Proof of work: each figure traced to the exact box or number it came from, plus the diagnostics the system ran.

A junior hands you a letter that started as an AI draft. Where does your time go?

Show the answer

Verifying every fact. The draft was the starting point; the finish line is the professional's.

Procurement wants insight into purchase orders and renewals. Route it.

Show the answer

Encoded and routine. With contracts and positions in a platform, procurement gains that insight.

Sources

The advice draws on interviews with the chief executive of a legal AI company, the head of a legal-analytics company, and the product chief of a professional software and data company.

Quiz

Join to take the quiz and see your score.

  1. A legal software vendor reports that 85% of legal professionals have some level of AI adoption at their firms. What is still yours to decide?

  2. Your team must summarize a long set of case documents for a client. A senior associate plans to write the first draft herself. What do you advise?

  3. A routine contract redraft comes back from the AI with its redlines marked. How much review before your team accepts it?

  4. Your audit team plans to use this routing tree as its complete route for a defined audit task. What should you tell them?

This is one lesson from The AI Decision Course.

$199 a year (founding price)

Blank card (PDF)The card as text, with a worked example

Membership adds every lesson and its card, the chat on your own case, your notes and progress saved, and a certificate per course.

Try it for 14 days. If it isn't useful, the full $199 comes back.

Paying with a company card? Ask your manager. Your receipt comes from Link, the checkout run by Stripe, and names the membership.

Made by Dr. Tony Hoang, host of The Artificial Intelligence Podcast.

Ask about this lesson

Which professional work AI drafts, and how much review each piece gets

    Free preview: 3 questions a day, answered from this lesson. Follow-ups remember your earlier questions on this visit. Don't paste confidential details.

    Ask about this lesson