Back to blog

50+ AI Prompts for Lawyers by Practice Area (2026)

Author: TheLawGPT Team|17 min|July 19, 2026
50+ AI Prompts for Lawyers by Practice Area (2026)
On this page

Most lawyers who try AI and walk away disappointed made the same mistake: they typed a one-line request and got a one-line-quality answer. The difference between "AI wasted my afternoon" and "AI saved me four hours" is almost entirely the prompt. This guide collects 50+ AI prompts for lawyers, organized by practice area, that you can copy, adapt, and start using today — whether you work in contract review, legal research, litigation, employment, family, or real estate law.

We also give you the framework behind the prompts, so you can write your own instead of hunting for templates every time. And we are honest about the part most prompt lists skip: no prompt, however well engineered, fixes a chatbot that invents case law. We cover where prompts stop working and what to do about it.

Want the output without the engineering? Ask TheLawGPT free at app.thelawgpt.com — these prompts work on any chatbot, but you still have to verify everything it says. TheLawGPT grounds every answer in verified legal databases and cites its sources. Free tier, no credit card.

Every effective legal prompt does five jobs. Miss one and the output degrades — usually into something generic, jurisdictionally wrong, or formatted uselessly.

ComponentPurposeExample
RolePuts the model in the right register and depth"Act as a California employment attorney"
JurisdictionPrevents the default blend of federal/general law"Apply New York law only"
ContextFacts, procedural posture, the document itself"My client is the subcontractor; the GC is withholding retainage"
Output formatControls structure so you can actually use it"Return a table of clause, risk level, and suggested revision"
Verification stepForces the model to show its work"Cite the specific statute or case for each point, and flag anything you are uncertain about"

A prompt using all five parts looks like this:

Act as a Texas real estate attorney. My client is the buyer in a residential
purchase; the seller failed to disclose foundation repairs. Under Texas law,
outline the available causes of action, the elements of each, and the notice
requirements. Return your answer as a bulleted list per cause of action,
cite the specific Texas statute or leading case for each, and flag any point
where the law is unsettled.

Two more habits that compound the value: iterate in follow-ups (treat the first answer as a draft you refine, not a verdict) and never paste identifying client information into a consumer chatbot — as of July 2026, the free and Plus tiers of ChatGPT still use conversations for model training by default, which is a confidentiality problem no prompt can solve. We break down exactly what happens to uploaded files in Is It Safe to Upload Legal Documents to ChatGPT?

With the framework in hand, here are the prompts.

Contract Review and Drafting Prompts

The highest-leverage use of AI in most practices. Upload or paste the agreement, then direct the analysis.

  • Full risk scan: "Act as a [state] contract attorney representing the [party]. Review this agreement and return a table of every clause that creates risk for my client, with columns for clause name, risk level (high/medium/low), why it is risky, and a suggested revision."
  • One-sided terms audit: "Identify every provision in this contract that is more favorable to [other party] than market standard for a [contract type], and explain what the balanced version of each provision would look like."
  • Missing protections: "List the protections a [buyer/seller/contractor/licensee] would normally expect in a [contract type] that are absent from this draft. Rank them by how much their absence hurts my client."
  • Indemnification deep-dive: "Explain the indemnification clause in this agreement in plain English: who indemnifies whom, for what, with what carve-outs, and whether the cap and survival period are market for a [industry] deal of this size."
  • Redraft a clause: "Redraft this [limitation of liability / non-compete / termination] clause to favor my client, the [party], while staying enforceable under [state] law. Provide the redline rationale in a sentence per change."
  • Defined-terms consistency: "Check this contract for defined terms that are used before they are defined, defined but never used, or used inconsistently, and list each with the section numbers."
  • Term sheet to draft: "Convert this term sheet into a first-draft [agreement type] under [state] law. Mark every provision where you made an assumption with [ASSUMPTION] so I can review each one."

Reviewing contracts every week? Upload one free at app.thelawgpt.com — instead of engineering the risk-scan prompt above, TheLawGPT runs the clause-by-clause review automatically, flags risky terms, and explains each in plain English. No credit card, results in minutes.

Research is where hallucination risk is highest, so these prompts build in verification. On a general chatbot, treat every citation as unverified until you confirm it in a real database — courts are fining lawyers who skip that step, as we document in our AI hallucination sanctions guide.

  • Elements of a claim: "Act as a legal researcher in [state]. List the elements of [cause of action], the leading case establishing each element, and the statute of limitations. Flag any element where courts in [state] are split."
  • Survey the landscape: "Give me an overview of how [state] courts have treated [legal issue] in the last five years: the majority approach, any minority approach, and the direction of the trend. Identify the key cases by name so I can pull them."
  • Statutory breakdown: "Break down [statute citation] section by section in plain English: who it applies to, what it requires, exceptions, penalties, and any regulatory guidance interpreting it."
  • Distinguish a bad case: "Opposing counsel relies on [case name and holding]. My facts are [facts]. Suggest every credible ground for distinguishing that case, ordered from strongest to weakest."
  • Circuit or state split: "Is there a split of authority on [issue]? Describe each camp's reasoning, which jurisdictions follow it, and which side [my jurisdiction] is on or would likely join."
  • Legislative update: "What changes to [area of law] in [state] took effect in the last 18 months, and what compliance steps do they require from [client type]? Note the effective date of each."
  • Research memo skeleton: "Draft the skeleton of a research memo on [question]: issue statement, brief answer, facts assumed, and the headings of the discussion section — no analysis yet, structure only."

Litigation and Motion Practice Prompts

AI cannot appear for you, but it can compress the drafting and preparation hours around every filing. For plaintiff-side workflows — medical records, demand letters, settlement valuation — we go deeper in AI for Personal Injury Lawyers.

  • Motion outline: "Act as a [state] litigator. Outline a motion to [dismiss/compel/exclude] based on these facts: [facts]. Give me the argument headings, the standard of review, and the authorities I need to find for each heading."
  • Discovery requests: "Draft a first set of [interrogatories/requests for production] for a [case type] where my client is the [party]. Focus on [key issues]. Keep each request narrow enough to survive an overbreadth objection."
  • Deposition prep: "Prepare a deposition outline for the opposing party's [role, e.g., project manager] in a [dispute type]. Organize by topic, and for each topic list the admissions I should try to lock in."
  • Anticipate opposition: "Here is my draft argument: [argument]. Act as opposing counsel and write the three strongest counterarguments, then suggest how I preempt each in my brief."
  • Timeline construction: "From these facts [or this document], build a chronological timeline of events with dates, actors, and the source for each entry. Flag gaps where I am missing evidence."
  • Demand letter: "Draft a pre-litigation demand letter on behalf of [client] to [recipient] regarding [dispute]. Firm but professional tone, deadline of [date], and reserve all rights. Keep it under one page."
  • Settlement analysis: "Given liability strengths [X], weaknesses [Y], provable damages of [amount], and venue [court], lay out a settlement valuation framework: best case, worst case, likely verdict range, and costs of trying the case."

Prompts get you a draft — sources make it usable. TheLawGPT searches millions of decisions and statutes and cites verified sources with every answer, so the checking step shrinks from hours to minutes. Free tier available.

Employment Law Prompts

Useful on both the counseling side and the dispute side.

  • Termination risk check: "Act as a [state] employment lawyer advising the employer. We plan to terminate an employee who [protected characteristic or protected activity, e.g., recently filed a workers' comp claim]. Walk me through the retaliation and discrimination exposure and the documentation we need before acting."
  • Handbook audit: "Review this employee handbook section for compliance with current [state] law. List each provision that is outdated, unlawful, or missing, and cite the controlling statute or regulation."
  • Non-compete enforceability: "Assess whether this non-compete is enforceable under [state] law as of 2026: duration, geography, scope, and consideration. Note recent statutory changes affecting enforceability."
  • Wage-and-hour exposure: "My client classifies [role] as exempt under [exemption]. Based on these duties [duties], analyze whether the classification survives scrutiny under [state] and federal law, and quantify the exposure formula if it does not."
  • Severance agreement: "Draft a severance agreement for a [state] employee at [level], including release of claims, [X weeks] severance, non-disparagement, and the OWBPA language required for employees over 40."
  • Investigation plan: "An employee alleges [harassment/retaliation] by [role]. Give me a step-by-step internal investigation plan: who to interview and in what order, what documents to preserve, and the memo structure for findings."
  • Policy drafting: "Draft a [remote work / AI use / PTO] policy for a [state] employer with [N] employees, in plain English, flagging every point where state law gives us less flexibility than federal law."

Family Law Prompts

Family matters are fact-heavy and emotionally loaded — AI helps most with structure, checklists, and first drafts.

  • Separation agreement checklist: "List every issue a [state] separation agreement should address for a couple with [children/property/business interests], organized by category, with the questions I need to ask my client under each."
  • Support calculation walkthrough: "Explain how [state] calculates child support when the payor earns [income] and has [custody arrangement]. Show the formula inputs and what documentation each requires."
  • Property characterization: "Under [state] law, walk through whether each of these assets is marital or separate property: [asset list with acquisition facts]. Explain the tracing issues for any commingled asset."
  • Parenting plan draft: "Draft a parenting plan for parents living [distance] apart with children aged [ages], covering the regular schedule, holidays, decision-making, communication, and relocation. Note where [state] courts typically want more detail."
  • Client intake summary: "Turn these intake notes into an organized case summary: parties, marriage timeline, children, assets and debts, income, immediate issues, and open questions — with a flag on anything time-sensitive."
  • Explain to the client: "Rewrite this explanation of [legal concept, e.g., equitable distribution] for a client with no legal background, in under 200 words, without losing accuracy."

Real Estate Prompts

From purchase contracts to leases, real estate work is document review at volume.

  • Purchase agreement review: "Act as a [state] real estate attorney for the [buyer/seller]. Review this purchase agreement and flag every deadline, contingency, and default provision, with the practical consequence of missing each."
  • Lease red flags: "Review this commercial lease for a [tenant type] tenant. Identify the ten most tenant-unfavorable provisions and suggest the market-standard alternative for each."
  • Title issue triage: "This title commitment lists the following exceptions: [exceptions]. Explain what each means for my client's use and financing of the property, and which ones we should demand be cured before closing."
  • Easement analysis: "Explain the scope of this easement language: who benefits, what uses it permits, whether it runs with the land under [state] law, and what my client can and cannot build."
  • Closing checklist: "Build a closing checklist for a [transaction type] in [state]: every document, who prepares it, who signs it, and the order of operations from contract to recording."
  • Disclosure obligations: "What must a residential seller disclose in [state], what forms are mandatory, and what is the liability exposure for nondisclosure? Note any 2025–2026 changes."

Practice-area context, built in. With TheLawGPT's Cases feature, you set the matter context once — upload the documents, describe the situation — and every question you ask already knows the full picture. No re-pasting, no re-explaining. Start a case free at app.thelawgpt.com.

Corporate and SME Compliance Prompts

For in-house counsel and the lawyers who serve small and mid-sized businesses.

  • Entity comparison: "My client is starting a [business type] in [state] with [N] owners and plans to [raise capital / stay closely held]. Compare LLC, S-corp, and C-corp on liability, tax treatment, and investor expectations for this fact pattern."
  • Privacy compliance gap: "My client is a [industry] company with customers in [states/regions]. List the privacy laws that apply, the core obligations of each, and the gaps in this current privacy policy: [policy text]."
  • Vendor contract playbook: "Create a negotiation playbook for [company type] reviewing inbound vendor contracts: our standard positions on liability caps, indemnity, data protection, auto-renewal, and termination, with fallback positions for each."
  • Board minutes: "Draft board meeting minutes from these notes: [notes]. Formal register, resolutions clearly stated, and a list of any follow-up filings or consents the resolutions trigger."
  • Regulatory horizon scan: "What regulatory changes affecting [industry] take effect in the next 12 months at the federal level and in [states]? For each, state the compliance deadline and the first operational step."
  • Corporate cleanup: "My client has not maintained corporate formalities since [year]. List the steps to bring a [state] [entity type] back into good standing and the documents to ratify past actions."
  • Terms of service draft: "Draft terms of service for a [product type] serving [customer type], covering acceptable use, payment, IP ownership, disclaimers, liability limits, and dispute resolution under [state] law. Flag clauses that consumer protection law restricts."

Client Communication Prompts

Underrated and nearly risk-free: no citations to hallucinate, immediate time savings.

  • Plain-English translation: "Rewrite this clause/opinion for a client with no legal background. Keep it accurate, under 150 words, and end with the one decision the client needs to make."
  • Status update: "Draft a client status email for a [matter type]: what happened since the last update [events], what happens next [steps], what we need from the client [items], and realistic timing. Reassuring but honest tone."
  • Bad news delivery: "Help me draft an email telling a client that [adverse development]. Acknowledge the setback, explain the options that remain, and recommend a path — without overpromising."
  • Engagement letter: "Draft an engagement letter for [matter type] with a [flat fee / hourly / contingency] arrangement under [state] bar rules, including scope, exclusions, billing terms, and file retention."
  • FAQ sheet: "Create a one-page FAQ for clients going through [process, e.g., a residential closing]: the eight questions they always ask, answered in two to three sentences each."
  • Intake questionnaire: "Build an intake questionnaire for a prospective [matter type] client that gathers the facts I need to spot conflicts, assess the claim, and quote a fee — grouped by topic, no legalese."
  • Difficult conversation script: "My client insists on [unwise course of action]. Draft talking points that explain the risk candidly, document that I advised against it, and preserve the relationship."

Where AI Prompts for Lawyers Stop Working

Here is the part most prompt roundups leave out: a perfect prompt cannot fix an ungrounded model. General-purpose chatbots generate citations by pattern-matching what a citation looks like — which is why courts have now sanctioned lawyers in hundreds of U.S. decisions involving fabricated authorities, with fines as of 2026 reaching five figures per attorney. Prompting "only cite real cases" does not reliably help; the model does not know which of its outputs are real.

There is also the confidentiality problem. As of July 2026, consumer tiers of ChatGPT use conversations for training by default, and a 2026 Southern District of New York ruling (United States v. Heppner) held that exchanges with a consumer AI platform were neither privileged nor protected work product. ABA Formal Opinion 512 puts the burden squarely on the lawyer: competence, supervision, and confidentiality apply to AI exactly as they do to a junior associate. We covered the accuracy gap in depth in ChatGPT for Lawyers: Why It Falls Short (And What to Use Instead).

The practical rule:

  • Safe on any AI: structure, outlines, checklists, plain-English rewrites, brainstorming counterarguments — anything you fully verify by reading it.
  • Needs a grounded legal AI: legal research, citations, jurisdiction-specific rules, document review of confidential material.

This is exactly the gap TheLawGPT is built for — and it removes the prompt-engineering burden at the same time. Every answer is grounded in verified legal databases updated daily and cites its sources with clickable links; if no reliable source exists, it says so rather than inventing one. The jurisdiction context the framework forces you to spell out is built in. The Cases feature remembers your matter context, so you stop re-pasting facts into every conversation. Clause-by-clause document review and document generation live in the same workspace, so the contract prompts above become a single upload. And your documents and conversations are never used for training, which keeps client material out of consumer-AI territory. The payoff is measurable — see our breakdown in How Much Time Does AI Save Lawyers? (Real Numbers).

Frequently Asked Questions

Can lawyers use ChatGPT and other AI tools ethically?

Yes. ABA Formal Opinion 512 (July 2024) and subsequent state bar guidance permit generative AI use provided the lawyer maintains competence, supervises the output, and protects confidentiality. In practice: verify everything, and never put client-identifying information into a tool that trains on your inputs.

Less than you think, and it depends on the tool. On general chatbots, prompt quality is the main lever — hence the framework above. Purpose-built legal AI like TheLawGPT needs far less scaffolding because the legal role, jurisdiction awareness, and source citation are built in; a plain-language question gets a grounded answer. You can test the difference on the free tier — three questions a day at app.thelawgpt.com, no credit card, $19.99/month when you outgrow it.

Can I put client information in AI prompts?

Only in tools with contractual confidentiality and no training on your data. Consumer chatbot tiers generally use conversations for training by default as of July 2026. Anonymize facts on consumer tools, or use a legal-specific platform that never trains on your data.

Will AI-generated citations be accurate?

On general-purpose chatbots, assume no — courts have sanctioned lawyers in hundreds of decisions involving fabricated citations. Verify every authority in a real database, or use a legal AI that only cites from verified sources and links each citation.

What is the best AI for lawyers in 2026?

It depends on budget and use case. Enterprise research platforms cost hundreds per user per month. For solo and small-firm lawyers, TheLawGPT covers legal Q&A with verified citations, clause-by-clause document review, and document generation from $19.99/month as of July 2026, with a free tier of three questions per day to test it.

Do these prompts work on any AI model?

The structure transfers everywhere — role, jurisdiction, context, format, verification. Output quality varies by model and, more importantly, by grounding: the same prompt produces a researched, cited answer on a source-grounded legal AI and a plausible-sounding guess on a general chatbot.

Copy the Prompts — or Skip Them Entirely

Save this page, adapt the placeholders to your jurisdiction and practice, and you will get dramatically better output from any AI tool you use. But if you would rather not engineer prompts at all, use a tool where the engineering is already done: ask your first question free at app.thelawgpt.com — verified sources, jurisdiction-aware answers, document review and generation included. Attorneys who use AI well report saving 10+ hours a week; at $19.99/month after the free tier, that is the best-leveraged line item in your practice. No credit card required, cancel anytime.

This article is for informational purposes only and does not constitute legal advice.