Tool Boundary Cards
Students match legal research tasks to the tools or sources best suited for the work. They decide whether a task belongs with AI, a legal research database, an official source, an annotated source, a human lawyer, or some combination, and explain where the boundary is.
Purpose
This exercise helps participants practice matching the research task to the right tool, source, or human judgment step.
Participants receive two sets of cards: task cards and tool/source cards. The task cards describe common legal research or legal work moments. The tool/source cards include options such as general AI, legal-specific AI, traditional legal research databases, annotated codes, official sources, court websites, secondary sources, and lawyer judgment. Participants match each task to the best tool or source and explain why.
The point is not to sort tasks into “AI” or “not AI.” The point is to help participants see that responsible AI use is task-specific. AI may be useful for orientation, vocabulary, issue spotting, search strategy, synthesis, or drafting support. But other tasks require official sources, verified authority, source hierarchy, confidentiality review, or professional judgment.
Learning Goals
By the end of this exercise, students should be able to:
- Identify the tool, source, or method best suited to a specific legal research task.
- Distinguish tasks appropriate for AI orientation from tasks requiring verified legal authority.
- Explain when traditional sources, official sources, annotated sources, or court materials are necessary.
- Recognize tasks that require lawyer judgment rather than tool output.
- Place AI within a broader research workflow instead of treating it as an all-purpose answer engine.
- Defend a tool choice in one or two sentences.
Materials
- Task cards describing legal research, drafting, verification, and professional judgment tasks.
- Tool/source cards naming available options.
- Optional boundary labels:
- AI-first.
- AI-assisted.
- Traditional-first.
- Official-source required.
- Human judgment required.
- Not enough information.
- Optional timer, whiteboard, or shared digital board.
- Optional instructor materials: model pairings, debrief prompts, and “hard call” cards for advanced discussion.
- Sample task cards:
- Get oriented to an unfamiliar doctrine.
- Generate search terms for a traditional database.
- Find controlling cases in a specific jurisdiction.
- Verify that a case is still good law.
- Check whether a statute is current.
- Determine whether a pending bill has changed the law.
- Find a local court’s standing order on AI disclosure.
- Compare two states’ statutory schemes.
- Locate cases interpreting a specific statutory phrase.
- Summarize five uploaded cases.
- Create a timeline from a set of client documents.
- Draft a first version of a client update.
- Decide whether a client should settle.
- Choose which argument to lead with in a motion.
- Determine whether confidential client facts can be entered into a tool.
- Sample tool/source cards:
- General-purpose AI tool.
- Legal-specific AI tool.
- Westlaw, Lexis, Bloomberg, or similar legal research database.
- Annotated code.
- Official code or government source.
- eCFR or agency website.
- Court website or local rules page.
- Secondary source.
- Citator.
- Client file or record.
- Internal firm knowledge system.
- Supervising attorney or lawyer judgment.
Setup
This exercise works best after participants have been introduced to the idea that different legal tasks create different risks.
Before beginning, participants should understand that no tool is universally appropriate. A tool that is useful for generating vocabulary may be inadequate for verifying authority. A tool that is helpful for drafting structure may be inappropriate for confidential facts. A tool that produces a plausible legal summary may still require traditional verification before the result can be used.
The instructor should frame the exercise around boundaries. The question is not simply “Can AI help?” The better question is: “What part of this task can AI help with, where does that help stop, and what source or human judgment step comes next?”
Run of Show
Total, approximately 30–45 minutes
- Frame the task Explain that groups will match legal work tasks to the tool, source, or judgment step best suited for the job. Their job is to identify both the possible use of AI and the boundary of that use.
- Distribute cards Give each group a set of task cards and tool/source cards. Groups may match one task to more than one source if the task requires a workflow rather than a single answer.
- Sort the easy cards first Groups begin with the obvious matches. For example, a citator is the right tool for checking whether a case is still good law; an official source is needed to confirm current statutory or regulatory text; lawyer judgment is needed for strategy and client advice.
- Debate the hard cards Groups then work through ambiguous tasks. These are the most useful cards. For example, AI may help generate search terms, but a legal research database may be needed to find and verify cases. AI may help draft a client update, but a lawyer must verify the law, adjust tone, and decide what advice to give.
- Add boundary labels For each task, groups label the task as AI-first, AI-assisted, traditional-first, official-source required, human judgment required, or not enough information.
- Defend the match Groups choose two or three cards and explain their match in one sentence. The explanation should identify why the tool fits and what still must be verified or decided by a lawyer.
- Debrief Class discussion focuses on disagreement. Which tasks were easy to place? Which ones required multiple tools? Where did AI help? Where did the group draw the boundary?
- Submission or reflection Participants submit a photo of their card sort or a short response identifying one task where AI helps, one task where AI is the wrong primary tool, and one task where lawyer judgment controls.
Student Instructions
Your group has two sets of cards. Task cards describe legal research, drafting, verification, or judgment tasks. Tool/source cards describe possible tools, sources, or decision-makers.
For each task card, choose the tool, source, or method that best fits the work. You may choose more than one if the task requires a workflow.
Then label the task:
- AI-first, AI is useful as the first step, but the output still needs review.
- AI-assisted, AI can help with part of the task, but another source or method is central.
- Traditional-first, a legal research database, citator, secondary source, or other traditional method should come first.
- Official-source required, the answer depends on current official text, court rules, agency materials, or government sources.
- Human judgment required, the task cannot be delegated to a tool because it requires legal strategy, client counseling, ethical judgment, or professional responsibility.
- Not enough information, you need more facts, context, jurisdiction, tool information, or client constraints before choosing.
For each match, be ready to explain:
- Why this tool or source fits the task.
- What AI can help with, if anything.
- What must still be verified.
- Where lawyer judgment is required.
Instructor Notes
The teaching value of this exercise comes from disagreement and boundary drawing.
Participants may initially treat the exercise as a simple sorting game: AI or no AI. Push them toward more precise answers. Many tasks are hybrid. AI may help with orientation, but not authority. AI may help with structure, but not strategy. AI may help summarize a source, but not verify whether the source is current or controlling.
The best debrief questions are:
- Where did your group disagree?
- Which task looked AI-friendly at first but required a traditional source?
- Which task looked traditional at first but could benefit from AI support?
- Which task required more information before you could choose a tool?
- Where did lawyer judgment enter the workflow?
Common issues to watch for:
- Participants may overuse AI for authority verification.
- Participants may underuse AI for search strategy, organization, or drafting support.
- Participants may forget official sources for current statutes, regulations, agency materials, or court rules.
- Participants may treat legal-specific AI as automatically verified.
- Participants may fail to distinguish between finding authority and relying on authority.
- Participants may assign strategic or ethical decisions to a tool.
Strong participant work usually shows:
- A clear match between task and tool.
- Recognition that many tasks require multiple steps.
- Attention to source hierarchy and currentness.
- A verification step before reliance.
- A distinction between tool output and lawyer judgment.
- A short explanation that connects the tool choice to the risk of the task.
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!