ResearchBeginner

Case Law Research Assistant

At the start of any legal research project, to build a foundation of relevant authority before diving into databases.

LitigationCorporateEmployment Law

Case law research is the bedrock of every brief, motion, and client opinion, but the first hours are often the slowest: framing the issue, surfacing the leading authorities, and separating the cases that help from the ones that hurt. The quality of that initial map shapes everything downstream, and starting from a blank Westlaw search wastes billable time better spent on analysis and strategy.

Claude is a capable research starting point. Give it the jurisdiction, the precise legal question, your favorable position, and the controlling facts, and it returns a structured memo identifying candidate leading cases, summarizing holdings and reasoning, flagging circuit splits, and proposing search terms to run in a primary database. It accelerates issue-spotting and framing so you arrive at Westlaw or Lexis already knowing what to look for.

Treat the result as a research roadmap, not authority you can cite. Claude can hallucinate case names, citations, and holdings, so every case it returns must be pulled, read, and Shepardized or KeyCited in a primary source before it touches a filing. The memo is a drafting aid for the attorney to verify and refine, not legal advice and not a substitute for independent research.

The Prompt

Research case law on the following legal issue:

Jurisdiction: [FEDERAL / STATE — specify circuit or state]
Legal issue: [DESCRIBE THE SPECIFIC LEGAL QUESTION]
Favorable position: [WHAT OUTCOME DO YOU WANT TO SUPPORT]
Relevant facts: [KEY FACTS THAT SHOULD MATCH THE CASES]
Time period: [e.g., last 10 years, or any landmark cases]

Provide:
1. The leading cases on this issue (3-5 cases), with full Bluebook citations
2. For each case: facts, holding, and key reasoning
3. How each case supports or undermines the favorable position
4. Any circuit splits or conflicting authority
5. The current trend in the law on this issue
6. Suggested search terms for Westlaw/Lexis to verify these citations

IMPORTANT: Flag any case you are not fully confident exists. I will verify all citations independently.

Example Output

A structured research memo with 3-5 case summaries, Bluebook citations, and analysis of how each applies to your facts.

Illustrative example — names, figures, and facts are fictional.

CASE LAW RESEARCH MEMO (DRAFT — VERIFY ALL CITATIONS)
Issue: Enforceability of a 12-month employee non-solicitation covenant absent a geographic limit (Ninth Circuit / California).
Favorable position: Covenant is unenforceable as to our client.

LEADING AUTHORITIES (illustrative placeholders — confirm in Westlaw/Lexis):

1. Marlowe Tech, Inc. v. Pace, No. 25-CV-1182 (N.D. Cal. 2025)
   Facts: Departing engineer solicited three former colleagues; agreement lacked geographic scope.
   Holding: Employee non-solicitation void under Cal. Bus. & Prof. Code analog; not a protectable interest.
   Effect: Strongly supports our position.

2. Veritas Sys. v. Okafor, 412 F. Supp. 4th 559 (N.D. Cal. 2024)
   Holding: Distinguished customer non-solicits from employee non-solicits; latter disfavored.
   Effect: Supports, but factually narrower.

3. Brennan v. Halcyon Labs, 88 Cal. App. 6th 1140 (2023)
   Holding: Upheld a narrowly tailored trade-secret-based restriction.
   Effect: UNFAVORABLE — opponent will rely on this; distinguish on trade-secret facts.

TREND: Post-2023 California decisions increasingly void standalone employee non-solicits.

SUGGESTED SEARCHES: "employee non-solicitation" /p enforceab! /s "protectable interest"; KeyCite Brennan for negative treatment.

CAUTION: Confirm each case exists and remains good law before citing.

Tips

•Always verify Claude's citations in Westlaw or Lexis — AI can hallucinate case names.
•Ask Claude to distinguish unfavorable cases rather than ignoring them.
•Specify the court level (Supreme Court, circuit, district) to get the most relevant authority.

Frequently Asked Questions

When should I use Claude for case law research?

Use it at the outset of a research project to frame the issue and build a roadmap of likely authorities before you open Westlaw or Lexis. Provide the jurisdiction, the precise legal question, your desired outcome, and the controlling facts. It is most valuable for issue-spotting and orientation, not for producing a finished, citable list of authorities.

Can I rely on the cases Claude returns without checking them?

No. AI models can fabricate plausible-looking case names, citations, and holdings. Every authority must be independently pulled, read in full, and validated with Shepard's or KeyCite in a primary source like Westlaw or Lexis before you cite it. Sanctions have followed lawyers who filed AI-generated citations without verification.

How do I get the most useful research output?

Be specific. Name the exact court or circuit, state the legal question narrowly, identify your favorable position, and list the facts the cases must match. Ask Claude to distinguish unfavorable authority rather than ignore it, and to suggest database search terms. Precise inputs yield a sharper, more useful research roadmap.

Is it ethical to use AI for legal research?

Yes, when supervised. ABA Formal Opinion 512 and the duty of competence (Model Rule 1.1) permit AI use provided the lawyer understands the tool's limits, verifies its output, and protects client confidentiality. The non-delegable obligation is independent verification: you remain responsible for the accuracy of every citation that reaches the court.

Get New Prompts Like This Every Week

Join the free Claude for Lawyers newsletter — weekly prompts, tutorials, and practice-specific guides.