AI prompts for legal research help US lawyers, paralegals, and law students structure the research phase of a project — issue spotting, case brief outlines, statutory summaries, memo drafting, comparative jurisdiction analysis — without doing the actual authoritative research. That distinction matters. AI is genuinely useful for scaffolding, brainstorming, and first-pass summaries. It is dangerous when treated as a substitute for Westlaw, Lexis, Bloomberg Law, or the primary sources themselves. Every prompt below is designed for the scaffolding job, not the authoritative-research job.
The critical warning that applies to every prompt on this page: AI hallucinates case citations. Every mainstream AI model — GPT, Claude, Gemini, Llama — has been shown to invent case names, docket numbers, holdings, and quotations that sound authoritative but do not exist. Multiple US lawyers have been sanctioned by federal courts (Mata v. Avianca, Park v. Kim, Rider v. Coffee-Enders) for filing briefs containing AI-hallucinated cases. Every case cite, statutory citation, and quotation must be independently verified in Westlaw or LexisNexis before appearing in any brief, memo, or client communication.
The prompts below are written for junior associates, paralegals, and law students who understand the underlying doctrinal areas and are using AI to move faster on the mechanical parts of research — organizing an issue list, drafting a case-brief template, planning a research strategy, structuring a memo. Every prompt starts with 'Act as a US legal researcher.' so the model gives US-law framing. Every prompt should be followed by verification of every citation, every quotation, and every doctrinal claim in an authoritative source. If you cannot verify it in Westlaw, LexisNexis, Bloomberg Law, or the primary source, do not cite it.
AI prompts for legal research help US lawyers, paralegals, and law students structure the research phase of a project — issue spotting, case brief outlines, statutory summaries, memo drafting, comparative jurisdiction analysis — without doing the actual authoritative research. That distinction matters. AI is genuinely useful for scaffolding, brainstorming, and first-pass summaries. It is dangerous when treated as a substitute for Westlaw, Lexis, Bloomberg Law, or the primary sources themselves. Every prompt below is designed for the scaffolding job, not the authoritative-research job.
The critical warning that applies to every prompt on this page: AI hallucinates case citations. Every mainstream AI model — GPT, Claude, Gemini, Llama — has been shown to invent case names, docket numbers, holdings, and quotations that sound authoritative but do not exist. Multiple US lawyers have been sanctioned by federal courts (Mata v. Avianca, Park v. Kim, Rider v. Coffee-Enders) for filing briefs containing AI-hallucinated cases. Every case cite, statutory citation, and quotation must be independently verified in Westlaw or LexisNexis before appearing in any brief, memo, or client communication.
The prompts below are written for junior associates, paralegals, and law students who understand the underlying doctrinal areas and are using AI to move faster on the mechanical parts of research — organizing an issue list, drafting a case-brief template, planning a research strategy, structuring a memo. Every prompt starts with 'Act as a US legal researcher.' so the model gives US-law framing. Every prompt should be followed by verification of every citation, every quotation, and every doctrinal claim in an authoritative source. If you cannot verify it in Westlaw, LexisNexis, Bloomberg Law, or the primary source, do not cite it.
Guides, tips, and deep dives for this prompt category
Can you sell AI art? Who owns the copyright? What are the legal risks? Clear answers to every legal question about AI-generated images.
Read morePromptsCreate stunning Studio Ghibli-style AI art with 50 free prompts for ChatGPT. Magical landscapes, characters, food scenes, and cozy interiors in Miyazaki style.
Read moreCopy any prompt below, paste into ChatGPT, Claude, Gemini, or Copilot, and fill in the placeholders in [brackets].
Act as a US legal researcher. Read this fact pattern and produce an issue-spotting list in the order the issues would be analyzed: constitutional issues first (if any), then federal statutory, then state statutory, then common-law claims, then defenses, then procedural issues. For each issue, list the general area of law and a starting research question — not a legal conclusion. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo. Facts: [paste].
Act as a US legal researcher. Draft a blank case-brief template with sections in this order: (1) case name and citation, (2) court and year, (3) procedural posture, (4) facts, (5) issue(s), (6) holding, (7) reasoning, (8) dicta (labeled as such), (9) disposition, (10) subsequent history. Include instructions for what belongs in each section. Do NOT fill in the template with a specific case — I will fill it from the actual opinion. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Summarize the general operation of statute [statute number, e.g. 15 U.S.C. § 1125] in plain English: what conduct it regulates, who has standing to sue, the elements of a claim, common defenses, and available remedies. Note that this summary is a starting point — the exact current text of the statute must be pulled from the U.S. Code (or state code) directly, and case law interpreting each element must be researched separately. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Draft a jurisdiction-specific research plan for [state] on this issue: [describe issue]. Include: (1) the primary state statute(s) to check, (2) the state supreme court and intermediate appellate court decisions to search for, (3) key search terms to use in Westlaw or Lexis, (4) relevant secondary sources (state practice guides, ALR annotations, treatises specific to the state), (5) any recent legislative activity to check on the state legislature's website. Deliver the plan as a checklist. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Design a secondary-source search strategy for this issue: [describe]. List: (1) the leading treatises to check (Wright & Miller for federal procedure, Corbin on Contracts, Prosser and Keeton on Torts, etc.), (2) top-tier law review articles to search for by topic in Westlaw or HeinOnline, (3) ALR annotations, (4) practice guides (BNA, Lexis Practice Advisor), (5) restatements. Note which sources are most authoritative in [jurisdiction] courts. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Produce a cite-checker outline for a brief. Steps: (1) pull every citation, (2) verify each cite exists in Westlaw or LexisNexis by clicking through, (3) confirm the case name, court, year, and reporter volume match, (4) run KeyCite / Shepard's on every case and note the treatment (positive, negative, questioned, overruled), (5) check pin cites against the actual page in the opinion, (6) verify statute language matches the current version, (7) verify quotations are verbatim, (8) flag any citation that cannot be found in an authoritative database as suspected AI hallucination. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Draft a memo structure with these sections: (1) Question Presented (one sentence, incorporates the legal question and the key facts), (2) Brief Answer (yes/no/probably plus one sentence of reasoning), (3) Facts (chronological, only facts material to the legal analysis), (4) Discussion (subheadings by issue, IRAC or CREAC for each), (5) Conclusion. Include a note about when to add a Statement of Applicable Law section. Do NOT fill in the memo — I will fill it from real research. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Draft a legislative history research plan for [statute or bill]. Include: (1) the Congressional record citations to check (House and Senate reports, committee reports, floor debates in the Congressional Record), (2) the bill's original text and each subsequent amendment, (3) the enacted public law number and Statutes at Large citation, (4) how to find hearings testimony (Congress.gov, ProQuest Legislative Insight), (5) presidential signing statements if any. Note the hierarchy of authority for legislative history in [jurisdiction]. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Draft a regulatory research plan for [federal agency, e.g. SEC / FTC / EPA / OSHA] on this issue: [describe]. Include: (1) the CFR title and part to review, (2) how to check the Federal Register for recent amendments and proposed rules, (3) agency guidance documents, no-action letters, and interpretive releases, (4) enforcement actions or settlements, (5) any pending petitions or rulemakings on the agency's docket. Note the deference standard (Chevron / Skidmore / Auer, as applicable and as modified by recent Supreme Court decisions). AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Draft a comparative-jurisdiction research plan for [issue] across [list states, e.g. California, New York, Texas, Florida, Illinois]. For each state, list: (1) the controlling statute, (2) the leading state supreme court decision, (3) the current majority-rule vs minority-rule position, (4) any recent legislative activity, (5) relevant state-specific secondary sources. Produce a comparison table skeleton I can fill in from real research. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Draft an argument outline for the [plaintiff / defendant / appellant / appellee] on this issue: [describe]. Structure using CREAC: Conclusion, Rule, Explanation of Rule (with citation slots for cases I will fill in), Application to our facts, Conclusion. Leave placeholders in [BRACKETS] for every case citation and quotation — I will fill them in from real research. Do NOT invent case names or quotations. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Act as a US legal researcher. Given my argument outline, generate the strongest counter-arguments the other side would raise. For each counter-argument, list: (1) the legal theory, (2) the type of authority that would support it (statute, case, secondary source), (3) how I would rebut it. Leave [BRACKETS] wherever a specific case or authority would go — do NOT invent case citations. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo. Argument: [paste].
Act as a US legal researcher. Prepare an oral argument prep list for [issue]. Produce the twelve most likely questions from a panel: three on the facts, three on the standard of review or procedural posture, three on the strongest doctrinal weakness in my position, three hypothetical questions extending the rule I am asking for to nearby fact patterns. For each, sketch a 30-second answer with placeholders for specific cases. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo. My argument: [paste].
Act as a US legal researcher. Read this record excerpt and extract the key facts material to a legal analysis of [issue]. Organize the facts chronologically and label each as (1) undisputed, (2) disputed but supported by admissible evidence in the record, (3) alleged but not supported in this excerpt. Do not invent facts or infer facts not present. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo. Record: [paste].
Understanding the building blocks lets you adapt any prompt to your own creative direction.
Tell the AI who the output is for and what real workplace situation it should support.
Act as a federal program analyst preparing a plain-language memo for agency leadership.Name the exact deliverable: email, memo, checklist, SOP, meeting recap, training note, or status update.
Format the answer as a one-page briefing with bullets, risks, and next actions.Specify whether the output should sound official, executive-ready, plain-language, or employee-friendly.
Use a professional, neutral, public-sector tone suitable for a US agency audience.For government, HR, finance, healthcare, legal, and compliance workflows, accuracy guardrails matter more than clever wording.
Use only the facts below, flag assumptions, and include a section for items that need verification.Ask the model to surface uncertainty so the user can verify sensitive or official information before using it.
Before finalizing, list compliance risks, missing details, and any claims that need human review.Tested on this prompt category as of mid-2026. Ratings reflect quality for AI Prompts for Legal Research specifically.
| Model | Best for | Rating |
|---|---|---|
| ChatGPT (GPT-4o / GPT-5) | Everyday drafting and summaries | |
| Claude Sonnet 4.5 | Long documents and policy | |
| Gemini 2.5 Pro | Grounded in Google workspace | |
| Copilot (M365) | Office 365 integration | |
| Perplexity | Answers with citations |
Ratings reflect suitability for this category. Free tiers available on all listed models. Last tested May 2026 by PromptSpace editors.
Yes for scaffolding (issue spotting, memo structure, research plans, brainstorming counter-arguments). No for authoritative substance (case citations, quotations, holdings, statute language, current-treatment analysis). Every citation must be independently verified in Westlaw, LexisNexis, Bloomberg Law, or the primary source. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Purpose-built legal AI tools (Westlaw Precision AI, Lexis+ AI, CoCounsel by Thomson Reuters, Harvey) hallucinate less because they retrieve from authoritative databases. General-purpose models (GPT, Claude, Gemini) hallucinate frequently on case law. Even with legal AI tools, every citation must still be verified. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Varies by court. Multiple federal district judges have entered standing orders requiring disclosure of generative AI use in any filed document, with specific certifications that all citations have been human-verified. The Fifth Circuit has a proposed rule. Some state courts have similar requirements. Check your court's standing orders, local rules, and your bar's ethics guidance before filing. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Not without confirming your firm's AI policy and your client engagement letter. Pasting privileged or client-identifying information into a public model can waive privilege, breach confidentiality, and violate your ethics rules. Use only AI tools that have been vetted by your firm for confidentiality and data-handling — typically enterprise deployments with contractual guarantees that inputs are not used for training. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Notify opposing counsel and the court immediately, file a corrected brief, and disclose the AI use if you have not already. Lawyers who have promptly self-corrected have generally received lighter sanctions than those who tried to defend the hallucinated cites. This is malpractice territory — consult your bar counsel and your firm's general counsel. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Learn the basics of creating stunning AI-generated images using prompts from our library.
GuideDiscover the secrets to crafting prompts that produce consistent, high-quality results.
CollectionCopy-paste 100 tested Midjourney v6 prompts: portraits, cinematic, fantasy, product shots & more. Free, updated for 2026 - instant results.
Social MediaCreate scroll-stopping Instagram content with these AI image prompts designed for Reels, Stories, and posts.
Browse our full library of ai prompts for legal research — all free, copy-paste ready, no signup.
Or use our AI Prompt Generator to create custom prompts for your exact style in seconds.
AI is good at explaining doctrines in plain English, generating issue-spotting checklists from a fact pattern, drafting the structural outline of a memo or brief, brainstorming counter-arguments, listing likely questions from an oral argument panel, and summarizing what a section of a statute appears to say. These are 'first-draft' tasks where being 80% right and easy to correct is more valuable than being 100% right at ten times the cost.
AI is also good at comparative overviews — 'how do these five states approach X issue in general terms' — as a starting point for real research. It is good at drafting research plans (what to look for, in what order, in which databases) that a paralegal can then execute. Treat AI output as a research plan and a scaffold, never as the source of doctrinal statements you would cite. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
AI is bad at everything that requires knowing whether a specific proposition is currently the law. It does not know whether the case it just cited was overruled, distinguished, or heavily criticized. It does not know whether the statute it just summarized was amended last month. It does not have reliable access to the underlying reporter volumes. It confabulates when its training data does not cover the specific holding it is stating, and it will not tell you when that happens — it will present a fabricated case with the same confidence as a real one.
AI is also bad at jurisdiction-specific nuance. It will summarize 'the majority rule' or 'the general rule' when the answer that matters is what your specific federal circuit or state supreme court says. It will apply the wrong version of a statute if the training data lags the current version. It will miss recent Supreme Court reversals, circuit splits, and pending appeals. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo, and check the case's current treatment (KeyCite, Shepard's) to confirm it is still good law.
Rule one: never cite a case, statute, regulation, or secondary source in a filing based solely on AI output. Every citation must be independently pulled from Westlaw, LexisNexis, Bloomberg Law, Fastcase, or the primary source, and the actual language must be read in the actual document. If the AI's quotation does not appear verbatim in the source you retrieved, do not use the quotation. Assume any specific pin cite from an AI is wrong until confirmed.
Rule two: use AI for structure, not substance. Ask AI to draft the outline of a memo, the structure of an argument, the list of issues to research, the categories of damages to consider. Then fill the substance from real research. Rule three: disclose your AI use as required by your local rules and standing orders. Multiple federal judges have entered standing orders requiring disclosure of generative-AI use in any filed document; some prohibit it entirely. Check your court's standing orders before submitting. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
A defensible workflow: (1) use AI to spot issues from the fact pattern and generate a research checklist, (2) execute the research in Westlaw or Lexis, reading actual cases and statutes, (3) use AI to draft the memo structure (question presented, short answer, facts, discussion, conclusion), (4) fill the discussion section from your real research with real Bluebook citations, (5) run every citation through KeyCite or Shepard's to confirm the case is still good law and not overruled, (6) have a colleague or senior lawyer review before filing, and (7) disclose AI use per your local rules.
This workflow gets you the productivity benefit of AI (faster scaffolding, better first drafts, quicker issue spotting) without the malpractice and sanctions risk of trusting AI substance. It also matches the way most firms are now setting AI policy: 'use it for structure and brainstorming, verify all citations, disclose per court rules, do not paste privileged or client-identifying information into public tools.' AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Yes for scaffolding (issue spotting, memo structure, research plans, brainstorming counter-arguments). No for authoritative substance (case citations, quotations, holdings, statute language, current-treatment analysis). Every citation must be independently verified in Westlaw, LexisNexis, Bloomberg Law, or the primary source. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Purpose-built legal AI tools (Westlaw Precision AI, Lexis+ AI, CoCounsel by Thomson Reuters, Harvey) hallucinate less because they retrieve from authoritative databases. General-purpose models (GPT, Claude, Gemini) hallucinate frequently on case law. Even with legal AI tools, every citation must still be verified. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Varies by court. Multiple federal district judges have entered standing orders requiring disclosure of generative AI use in any filed document, with specific certifications that all citations have been human-verified. The Fifth Circuit has a proposed rule. Some state courts have similar requirements. Check your court's standing orders, local rules, and your bar's ethics guidance before filing. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Not without confirming your firm's AI policy and your client engagement letter. Pasting privileged or client-identifying information into a public model can waive privilege, breach confidentiality, and violate your ethics rules. Use only AI tools that have been vetted by your firm for confidentiality and data-handling — typically enterprise deployments with contractual guarantees that inputs are not used for training. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.
Notify opposing counsel and the court immediately, file a corrected brief, and disclose the AI use if you have not already. Lawyers who have promptly self-corrected have generally received lighter sanctions than those who tried to defend the hallucinated cites. This is malpractice territory — consult your bar counsel and your firm's general counsel. AI hallucinates case citations — verify every case in Westlaw or LexisNexis before citing in any brief or memo.