Responsible AI in Legal Practice
Seven exercises for exploring prompt strategy, legal reasoning, and AI-assisted research.
Each exercise below builds on core prompting skills for legal research. Select a scenario to open a step-by-step guide — from initial query construction through jurisdiction-specific analysis and output formatting. Exercises 2 through 5 form a connected sequence; the others stand alone.
Citation Extraction and Formatting
Verifying Citations from a Pro Se Filing
Facing a pro se litigant in a debt settlement action whose citation-heavy filings include suspect authority — checking each efficiently without a major research investment.
Basic Prompt Formatting
Structuring the Initial Query
A residential tenant discovers their landlord has been entering without notice. Comparing a plain-language query to a structured legal fact pattern to see how framing shapes results.
Refining Tone and Guidance
Adding Instruction to the Prompt
Returning to the tenant privacy question from Exercise 2 with added guidance language — directing the model on how to think, structure its response, and calibrate its tone.
Adding Detail to Prompts
Introducing Jurisdictional Specifics
With a basic legal framework in hand, refining the analysis by identifying controlling authority, applicable statutes, and unsettled law within a specific jurisdiction.
Boolean Terms and Connectors
Generating Westlaw Search Queries
Using the legal framework developed in prior exercises to generate Boolean search queries formatted for Westlaw — bridging AI-assisted research to traditional citator tools.
Additional Prompt Options
Follow-Up Prompts for Deeper Analysis
A set of follow-up prompts for use when initial results lack pivot points, or when a more structured analytical output — charts, outlines, burden analysis — would be useful.
Additional Questions
Practice Across Areas of Law
Eight fact patterns spanning environmental, employment, advertising, constitutional, IP, health, estates, and securities law — for applying the prompting patterns from earlier exercises.