The Workflow Map

Purpose

This exercise helps students see legal research as a workflow, not a single search task.

Students map the steps they would take to move from an initial legal question to a usable work product. Then they decide where AI could help, where traditional research is still required, where a lawyer must exercise judgment, and where a verification gate belongs.

The point is not to create one perfect workflow. The point is to make students notice that AI does not enter “legal research” all at once. It may help at one step, create risk at another, and be irrelevant to a third. The exercise also helps separate legal research pain points from administrative or workflow pain points, which can lead to different tool choices and different supervision needs.

Learning Goals

By the end of this exercise, students should be able to:

1. Break a legal research task into discrete workflow steps.
2. Identify where AI may add value through orientation, organization, synthesis, drafting support, or issue spotting.
3. Distinguish between steps that can be AI-assisted and steps that require traditional research, legal authority, or independent professional judgment.
4. Place verification gates at points where factual, legal, or citation accuracy must be checked before moving forward.
5. Separate legal research pain points from administrative, organizational, or communication pain points.
6. Explain a workflow choice in language usable with a supervising attorney or training audience.

Materials

• Large notepad paper, whiteboard space, sticky notes, or a shared digital board.
• A sample legal research scenario or practice task.
• Markers or sticky-note colors for categories such as: AI-assisted step, traditional research step, lawyer judgment step, verification gate, and pain point.
• Optional workflow labels: research, analyze, validate, draft, revise, communicate.
• Optional instructor materials: sample completed workflow map, list of common verification gates, and debrief questions.

Setup

This exercise works best after students have already seen examples of AI-assisted research, drafting, or verification.

Before beginning, students should understand that AI is not simply “used” or “not used” for a whole assignment. Its usefulness depends on the stage of the work. It may be helpful for orientation, synthesis, search strategy, summarization, comparison, or drafting structure. It may be inappropriate for final authority checking, strategic decisions, client advice, or filing-ready work without independent review.

The instructor should frame the exercise around a realistic professional task. Students are acting as a team of associates, librarians, clinic students, or legal professionals asked to design a responsible workflow before beginning the work.

Run of Show

Total, approximately 30–45 minutes

1. Frame the task Give students a legal research scenario or ask them to use a familiar research task. Explain that they are not solving the legal issue. They are mapping how the work should happen.
2. Build the workflow Students map the steps from initial question to final deliverable. Encourage them to include all work, not just database searching: intake, issue framing, source selection, search strategy, reading, note-taking, authority checking, drafting, partner communication, and final review.
3. Mark AI-assisted steps Students identify which steps might benefit from AI, issue spotting, generating search terms, organizing notes, summarizing known materials, comparing sources, building an outline, or drafting a first-pass structure.
4. Mark traditional steps Students identify which steps must remain traditional or source-grounded, reading statutes, verifying cases, checking currentness, using citators, consulting authoritative secondary sources, confirming procedural rules, and reviewing the actual text of authority.
5. Add lawyer-judgment steps Students mark where the lawyer must stay in control, defining the research question, choosing the client strategy, deciding which authorities matter, weighing risk, communicating advice, and signing off on final work product.
6. Place verification gates Students add verification gates between steps. A gate should appear anywhere a bad assumption could travel forward: after AI issue spotting, after AI-generated authority lists, before drafting, before advising a client, before citing authority, and before sending or filing anything.
7. Identify pain points Students mark pain points and classify each one: is this a legal research pain point, a source-access pain point, an administrative pain point, a communication pain point, or a documentation pain point?
8. Debrief Groups briefly explain one AI-assisted step, one traditional step, one lawyer-judgment step, and one verification gate. The class discussion focuses on why those choices differ across workflows.
9. Submission or reflection Students submit a photo of the workflow map or a short paragraph explaining one workflow change they would make after seeing the map.

Student Instructions

Instructor Notes

The teaching value of this exercise comes from making invisible workflow decisions visible.

Students may initially map only the obvious research steps: search, read cases, cite-check, draft. Push them to include the surrounding work: clarifying the assignment, deciding what the client actually needs, choosing sources, tracking time, organizing notes, communicating uncertainty, and documenting what was verified.

The exercise also helps students see that not every pain point is a legal research pain point. A student may think they need a better AI research tool when the real problem is note organization, version control, task handoff, source access, or unclear supervisor expectations.

Common issues to watch for:

• Students may put AI too late in the process, using it only for drafting.
• Students may put AI too early without a confidentiality or framing check.
• Students may forget to verify AI-generated authorities before relying on them.
• Students may treat verification as one final step instead of a recurring gate.
• Students may overlook administrative pain points such as tracking sources, documenting decisions, or communicating uncertainty.
• Students may fail to distinguish “AI can help organize this” from “AI can decide this.”

Strong student work usually shows:

• A clear sequence from question intake to final communication.
• Specific AI use cases tied to particular workflow steps.
• Traditional research steps where authority, currentness, and source hierarchy matter.
• Multiple verification gates, not just one final review.
• Identification of non-research pain points.
• Recognition that lawyer judgment controls framing, strategy, reliance, and final advice.

Adaptation

Here are some suggestions on how to adapt this in other teaching contexts. The adaptations below have been limited to firm trainings and asynchronous suggestions, but there are other possible adaptations for workshops, seminars, and other contexts. Be creative!