AI Prompts for Lawyers: A 5-Part Method and 10 Ready-to-Use Legal Prompts

Good AI prompts for lawyers follow a precise structure: a role for the AI, the area of law and jurisdiction, anonymized facts, the expected output format, and an explicit request to flag uncertainty. A vague prompt gets generic or wrong answers. A structured one turns the tool into a reliable assistant, and it never contains confidential client information.

This guide gives you the method, then 10 ready-to-use prompts for ChatGPT, Perplexity, Gemini, and Microsoft Copilot, written for US practice. It closes with the ABA rules to follow and an overview of the legal AI tools available to US firms.

What are AI prompts for lawyers, and why does precision matter in law?

A prompt is the instruction you give a generative AI tool. In law, the gap between a vague request and a structured one changes everything: a loose prompt gets a generic, sometimes wrong answer, while a framed prompt steers the model toward output you can use. The risk is not theoretical. General-purpose models make factual errors on legal questions at a high rate, and specialized tools are not immune either.

Better tools do not make hallucinations disappear

In a 2024 benchmark by Stanford RegLab and HAI, general-purpose chatbots hallucinated on 58% to 82% of legal queries. A 2024 Stanford study cited in ABA Formal Opinion 512 found that the AI tools of leading legal research companies still hallucinated between 17% and 33% of the time. Checking every citation remains essential, whatever the tool.

Many lawyers use AI, few have a method

Adoption is already real in the US. The ABA's 2024 Legal Technology Survey Report found that 30% of responding lawyers used AI-based tools, up from 11% in 2023, and Clio's 2024 Legal Trends Report put AI use in some form at 79% of legal professionals. In the ABA survey, accuracy was the top concern, cited by 75% of respondents.

In other words, most lawyers have started using AI, but few know how to brief it. Under Opinion 512, competence includes a reasonable understanding of what an AI tool can and cannot do. That is where the prompt makes the difference.

Prompt engineering for lawyers: the 5-part structure of a legal prompt

Rather than copying prompts you do not understand, memorize this structure. It adapts to any task, any practice area, and any state.

The five parts

• 1. The role you give the AI: "You are a legal assistant specializing in [State] employment law."
• 2. The area of law and the jurisdiction: always state "[State] law" or "federal law", and name the court, such as "U.S. District Court for the [District]" or "the [County] trial court". Without it, the model may answer under another state's rules or blend federal and state law.
• 3. The facts, without the conclusion: describe the anonymized facts without hinting at the answer you expect, so you do not bias the analysis.
• 4. The output format: bullet points, a structured memo, a formal tone, a length limit. Stating the format you expect noticeably improves the answer.
• 5. The verification request: always end with "Flag any uncertainty. Cite no case or statute unless you are certain it exists, format every citation in Bluebook style, and mark each one [VERIFY]."

Why the verification step matters most

This last part is the most important. It does not eliminate hallucinations, but it shows you where to look first and keeps you in control. Every authority marked [VERIFY] still gets pulled in Westlaw or Lexis and checked with KeyCite or Shepard's before it goes anywhere.

For the broader picture of uses and limits, see our guide to ChatGPT for lawyers and our article on ChatGPT in legal practice: risks and best practices.

ChatGPT prompts for lawyers: 4 examples for US practice

ChatGPT excels at first drafts and rewriting. One absolute rule: use these prompts on anonymized or fictitious facts only. On consumer plans (Free, Plus, and Pro), OpenAI may train on your chats unless you turn off "Improve the model for everyone" in Data Controls. ChatGPT Business and Enterprise do not train on your data by default, but that setting alone does not discharge your duty of confidentiality.

Drafting prompts

• Matter summary (anonymized): "You are a legal assistant. From these anonymized facts [paste], write a summary of 10 lines maximum: parties (anonymized), nature of the dispute, claims, and legal issues. Flag any area of uncertainty."
• First draft of a demand letter (negligence): "Draft a demand letter for a negligence claim under [State] law, in a formal tone. Leave every element the attorney must complete in [brackets]. Invent no amount, date, or statute, and mark any legal reference [VERIFY]."

Intake and research prompts

• Intake call script (family law): "Create a list of questions for a first call with a prospective client going through a divorce in [State], to qualify the situation (marital property, children, whether both spouses agree on the divorce) without giving legal advice."
• Research starting point: "Explain the requirements for an uncontested divorce in [State] and the documents usually needed. Tell me if a rule has changed recently, give any statute citation in Bluebook format marked [VERIFY], and remind me to check the current statutes and court rules."

Information from prospective clients is protected under Rule 1.18, so an intake script collects only what you need. More in our guide to client intake automation.

Perplexity prompts for lawyers: sourced legal research

Perplexity cites its sources and lets you go back to the original documents. That makes it a good tool for monitoring a topic and for first-pass research, as long as you check every link: cited URLs can be wrong or point to pages that have changed. A lower error rate never removes the need to verify; it only lowers the cost of verifying.

Case law and procedure prompts

• Case law monitoring: "What are the main recent trends in [State] appellate decisions on damages for emotional distress? Cite your sources with links and dates, and give each case citation in Bluebook format."
• Procedure comparison: "Compare in a table a contested divorce and an uncontested divorce in [State]: typical duration, indicative cost, role of the judge, and required documents. Cite your sources."

Legislative tracking prompt

• New legislation: "Summarize the main recent legislative changes in [State] employment law on [topic, for example noncompete agreements]. Give the bill or statute numbers and the effective date of each change, with links to the official sources."

Treat every answer as a lead to read at the source, never as authority. Our guide to Perplexity for lawyers goes further.

Gemini and Copilot prompts: document analysis and hearing prep

Gemini (Google) and Microsoft 365 Copilot are especially useful for working on long documents inside the tools your firm already runs. One caution: for any confidential document, use only a business version your firm has vetted, such as Microsoft 365 Copilot within your Microsoft 365 plan or Gemini in a Google Workspace business plan, with written data terms. Never the consumer app.

Document analysis prompt

• Document review (anonymized): "Analyze this document [paste anonymized text]. Identify sensitive clauses, ambiguities, and points to verify under [State] law. Present the result as a table."

Hearing, deposition, and mediation prompts

• Hearing or deposition prep: "From this anonymized fact pattern, list 15 possible cross-examination questions and the angles opposing counsel is likely to take. Do not state any final legal conclusion."
• Mediation strategy: "Propose 3 negotiation scenarios for the mediation of a commercial contract dispute governed by [State] law, with the arguments and possible concessions for each side."

An employment law firm can use the same approach to draft letters quickly during a negotiation. See our guides to Gemini for lawyers and Microsoft Copilot for lawyers. For the content and visibility side, read our guide to law firm content marketing.

The 6 use cases where legal prompts pay off first

If you had to prioritize, here is where the pairing of a good prompt and the right tool adds the most value, without crossing ethical lines.

Research and drafting

• Case law research and monitoring with Perplexity or a legal AI platform, as our guide to AI legal research explains.
• First drafts of motions, memos, and letters, always reviewed line by line. See our guide to AI legal writing.
• Matter summaries built on anonymized data.

Clients, marketing, and contracts

• Client communication: rewriting complex memos in plain English.
• Marketing content and blog posts for your visibility, drafted with the safeguards in our guide to AI content for law firms and tied to AI search optimization for law firms.
• Contract review: spotting clauses and comparing versions, covered in our guide to AI contract review.

On that last point, a real estate lawyer can, for example, compare several lease offers side by side. To turn that productivity into growth, read our law firm SEO tips.

Risks and ethics: what ABA Formal Opinion 512 means for your prompts

This is the heart of the matter. No US jurisdiction bans lawyers from using AI, and the rules you already follow apply in full. ABA Formal Opinion 512, issued on July 29, 2024, is the ABA's first formal ethics opinion on generative AI. It ties AI use to your existing duties: competence, confidentiality, communication, candor, supervision, and reasonable fees.

Client information stays out of raw prompts

The non-negotiable point is client data. Rule 1.6(c) requires reasonable efforts to prevent the unauthorized disclosure of client information. Opinion 512 requires the client's informed consent before you enter information relating to the representation into a self-learning tool, and a boilerplate clause in an engagement letter is not enough.

State bars add their own layer. The State Bar of California's guidance (November 2023) asks lawyers to anonymize client details, and Florida's Ethics Opinion 24-1 (January 2024) recommends informed consent before sharing confidential information with a third-party AI tool. Check the guidance in each state where you practice.

Vendor terms and privacy laws

Pasting a client's facts into a third-party model discloses them to that provider. Before you adopt a tool, ask where data is stored, how long it is kept, who can access it, whether it trains models, and whether the vendor holds an independent security audit such as a SOC 2 report. State privacy laws such as the CCPA may also apply.

Responsibility stays with the lawyer

Responsibility does not move to the software. Under Rules 5.1 and 5.3, supervising lawyers set policies for everyone who uses AI, and in federal court a Rule 11 signature certifies that your legal contentions are warranted by existing law. In Mata v. Avianca (S.D.N.Y. 2023), two New York lawyers and their firm were fined $5,000 for a brief citing cases ChatGPT had invented.

The best practice remains a one-page internal AI policy, shared with the whole team: approved tools, approved uses, and the data that never goes into a prompt.

Legal AI tools: the US landscape

Beyond general assistants, the legal AI market is taking shape. Harvey mostly serves large firms and in-house teams: Allen & Overy, now A&O Shearman, announced a firmwide rollout in February 2023. Legal publishers ground their AI in their own content: CoCounsel from Thomson Reuters, the Lexis+ platform from LexisNexis, and vLex's Vincent AI. Spellbook works inside Microsoft Word for contract drafting.

Data terms come before performance

The deciding criterion is not raw performance but confidentiality. A "secure" or "enterprise-grade" label on a vendor's website is neither sufficient nor verifiable on its own. Opinion 512 expects you to understand a tool's terms of use and privacy policy before you rely on it, so ask for the written data terms and check how your data is processed before you sign.

Tools and confidentiality checkpoints

For a fuller comparison, see our roundup of legal AI tools and our guide to AI for lawyers. Building all of this into a real visibility and client acquisition strategy is our work as a marketing agency for lawyers.

Key takeaways

A good legal prompt rests on five parts: role, area of law and jurisdiction, anonymized facts, output format, and a verification request. The tool matters less than the method and your duty of confidentiality under Rule 1.6. By staying in control of every output, you turn AI into a real productivity gain without ever delegating your responsibility.

Mastering AI prompts for lawyers is no longer a gimmick: it is a competitive advantage, as long as client confidentiality comes first. We help firms build a complete strategy, from the prompt to online visibility. Tell us about your firm.