Best ChatGPT Prompts for Mortgage Brokers in 2026

Each prompt below is written as a self-contained instruction you paste into ChatGPT, Claude, or another capable model. Bracketed placeholders like [BORROWER NAME] or [LOAN PROGRAM] are yours to fill in before pasting. If the model's first draft is close but not perfect, use a follow-up: "Shorten this to three sentences" or "Make the tone warmer and less formal" will get you the revision in seconds.

The prompts in this guide are organized by workflow stage: explaining products, updating borrowers, checklists, rate-lock communications, scenario comparisons, realtor outreach, review requests, first-time-buyer education, objection handling, and internal guideline summaries. Within each section you will find one to three ready-to-use prompts plus notes on when to use them and what output to expect.

One structural tip that improves every prompt: lead with role context, then task, then constraints. "You are a mortgage loan officer. Draft a plain-English explanation of [X] for a borrower with no prior homebuying experience. Keep it under 200 words. Do not quote specific interest rates or APR figures — use placeholders instead." That pattern reliably outperforms a bare question and keeps the output within your compliance guardrails. For a deeper look at prompt structure, see how to write better prompts: 15 rules.

One of the highest-value uses of ChatGPT for mortgage brokers is translating loan program jargon into plain English a first-time buyer can actually understand. The key constraint: the prompt must not ask the model to quote live rates, APR figures, or pricing — those change daily and must come from your pricing system, not an AI. Use placeholders like [CURRENT RATE PLACEHOLDER] and fill them in manually.

Prompt — FHA vs. Conventional plain-English explainer: ``` You are a licensed mortgage loan officer writing educational content for borrowers. Explain the difference between an FHA loan and a conventional loan in plain English. Cover: minimum down payment, mortgage insurance requirements, credit score considerations, and which borrower profile tends to fit each program better. Do not quote specific interest rates or APR figures — use the placeholder [RATE WILL BE PROVIDED BY YOUR LOAN OFFICER] wherever a rate would normally appear. Keep the explanation under 300 words. Tone: friendly, clear, no jargon. ``` Use this when: a borrower asks "should I go FHA or conventional?" and you want a quick, accurate, compliant explainer to send alongside your personal recommendation. Always add your own rate context and a note that programs vary by lender.

Prompt — VA loan benefits explainer for a veteran borrower: ``` You are a mortgage specialist writing for a first-time veteran homebuyer. Explain the key benefits of a VA loan in plain English: no-down-payment option (up to conforming loan limits), no private mortgage insurance, VA funding fee structure (describe it conceptually, not as a specific dollar amount), and the Certificate of Eligibility process. Keep it under 250 words. End with a sentence encouraging the borrower to contact their loan officer to confirm eligibility for their specific situation. ``` Compliance note: the VA funding fee varies by service type, down payment amount, and whether it is a first or subsequent use. Never let AI output a specific fee dollar amount — describe it structurally, then point to VA.gov or your own fee schedule.

Keeping borrowers informed throughout a loan transaction is both a business practice and — at certain stages — a regulatory requirement under RESPA. AI is well-suited to drafting the non-regulated, courtesy communication: the "here's where we are" email that reduces inbound calls and builds confidence. Always send these as drafts through your LOS communication log so there is a record.

Prompt — Mid-process status update email: ``` You are a mortgage loan officer. Draft a short, professional email to a borrower updating them on the status of their loan application. Fill in the blanks with these details: Borrower name: [NAME]. Current stage: [e.g., "We have received your appraisal and are now in underwriting"]. Estimated next milestone: [e.g., "We expect a conditional approval within 3-5 business days"]. One action item the borrower needs to complete: [e.g., "Please upload your most recent pay stub to the portal"]. Tone: warm, reassuring, professional. Length: 4-6 sentences. Include a closing line reminding them you are available for questions. ``` This prompt produces a usable first draft in seconds. A processor or LO spends about 30 seconds personalizing it. Run it for every in-progress file and your communication touchpoints increase with no additional labor.

Prompt — Clear-to-close congratulations email: ``` You are a mortgage loan officer. Write a warm, enthusiastic clear-to-close email for a borrower named [NAME] who is purchasing a home at [CITY, STATE]. The closing date is [DATE]. Cover: what clear to close means in plain English, what happens next (scheduling closing, final walk-through reminder, bringing a cashier's check or wire for [PLACEHOLDER AMOUNT] — do not insert a specific dollar figure), and express genuine congratulations. Tone: celebratory but professional. Length: 150-200 words. ```

Building a tailored pre-qualification checklist for each borrower scenario — W-2 employee, self-employed, recent job-change, divorce decree needed — is a task ChatGPT handles well when you describe the scenario clearly. These outputs are internal tools and borrower-facing guides, not official disclosures, so they carry lower compliance risk than rate-specific communications.

Prompt — W-2 borrower pre-qual document checklist: ``` You are a mortgage loan officer creating a document checklist for a W-2 borrower applying for a conventional purchase loan. List every document typically needed for initial pre-qualification, organized into categories: Income Verification, Asset Verification, Identity and Credit, and Property (if already under contract). For each item, add a one-sentence note explaining why lenders need it. Format as a numbered checklist. Do not include anything specific to a particular lender's internal forms or overlays. ``` Output from this prompt makes an excellent borrower-facing PDF or a page on your website. Have your compliance team review the language once, then reuse the template for every W-2 file.

Prompt — Self-employed borrower checklist (additional items): ``` You are a mortgage loan officer. A borrower is self-employed and applying for a conventional mortgage. Generate the additional documentation items — beyond a standard W-2 checklist — that a self-employed borrower typically needs to provide. Include: business tax returns (specify years), personal tax returns (specify years), year-to-date profit and loss statement, business bank statements, and business license if applicable. For each item, note what the underwriter is typically looking for. Format as a numbered list. Do not quote specific income thresholds or DTI ratios — those vary by program and lender. ```

Rate locks are one of the most misunderstood concepts for first-time borrowers and a frequent source of disputes when deals extend or rates move. A clear, written explanation — sent at lock time — protects you and educates the borrower simultaneously. ChatGPT can produce a clean explainer in seconds; a human must add the actual lock terms (duration, expiration date, extension cost policy) before sending.

Prompt — Rate-lock confirmation explainer: ``` You are a mortgage loan officer. Draft a plain-English explanation of what a rate lock is and why it matters, to be sent to a borrower at the time their rate is locked. Cover: what a rate lock does (freezes the interest rate for a set period), what happens if the loan does not close before the lock expires, and what a rate lock does NOT protect against (e.g., changes in loan program, property appraisal issues, borrower qualification changes). Do not include specific rate figures or lock period durations — those will be added manually. Tone: educational, reassuring. Length: 150-200 words. ``` After the model produces this text, you add: "Your rate has been locked at [RATE] for [DURATION] days, expiring on [DATE]. Lock extension fees, if needed, are [POLICY]." That final section is not AI-generated; it comes from your pricing desk.

Prompt — Rate-lock expiration warning (file running long): ``` You are a mortgage loan officer. Draft a professional email to a borrower whose rate lock is approaching expiration. The tone should be calm and informative, not alarmist. Explain that the lock expiration date is approaching, that you are actively working to close on time, and that if an extension is needed you will discuss the process and any associated cost with them directly. Do not include specific dollar amounts for extension fees. Include a clear call to action asking the borrower to promptly return any outstanding documents. Length: 100-150 words. ```

When a borrower is weighing two loan options — a 30-year fixed vs. a 7/1 ARM, or a 15% down conventional vs. 20% down to avoid PMI — a structured pros/cons comparison helps them make a decision without requiring you to write a long memo from scratch. The critical rule: ask ChatGPT to structure the comparison without generating specific monthly payment figures or rate assumptions. Those come from your actual pricing.

Prompt — 30-year fixed vs. ARM comparison framework: ``` You are a mortgage loan officer creating an educational comparison for a borrower deciding between a 30-year fixed-rate mortgage and a 7/1 adjustable-rate mortgage (ARM). Build a structured pros/cons table for each product covering: payment predictability, initial rate environment (describe conceptually — do not quote specific rates), risk profile, who each product tends to suit best, and the break-even scenario where an ARM makes more sense than the fixed. Use placeholder language like "your loan officer will provide the specific rate comparison" wherever actual numbers would appear. Tone: balanced, educational. Format: two-column comparison table followed by a 2-paragraph summary. ``` This is an internal communication and educational tool, not a disclosure. When you present the comparison to the borrower, you attach your actual rate sheet numbers alongside it.

Prompt — PMI vs. 20% down decision framework: ``` You are a mortgage professional. A borrower has enough savings to either put 10% down (and pay PMI) or stretch to 20% down (and avoid PMI, but deplete their cash reserves). Write a structured pros-and-cons analysis of each approach. Cover: monthly cash flow impact (describe conceptually), reserve adequacy after close, PMI cancellation rules under the Homeowners Protection Act, opportunity cost of the larger down payment. Do not calculate specific monthly payment figures — use descriptive placeholders. End with a note that a loan officer can run the exact numbers once the rate and loan amount are confirmed. Length: 300-400 words. ```

Realtor relationships are the primary referral engine for most purchase-market mortgage brokers. ChatGPT can draft outreach emails, follow-up messages, and co-marketing proposals faster than any template library. The goal is to produce a message that sounds human, references the realtor's business context, and makes a clear value proposition — all in under 60 seconds of prompt work.

Prompt — First-touch email to a real estate agent: ``` You are a mortgage broker writing an initial outreach email to a real estate agent you have not worked with before. The agent's name is [AGENT NAME] and they specialize in [MARKET/NEIGHBORHOOD]. Your value proposition: fast pre-approvals (mention 24-hour turnaround), access to multiple lenders for competitive pricing, and responsive communication through the transaction. Write a professional, not salesy, email of 120-150 words. Open with a genuine reference to their market or recent activity — use a placeholder like "I noticed you have been active in [NEIGHBORHOOD] recently". Close with a low-pressure ask: a 15-minute phone call or coffee. ```

Prompt — Co-marketing event proposal to a realtor partner: ``` You are a mortgage broker writing to a real estate agent you have an existing referral relationship with. Propose a joint first-time homebuyer seminar. Cover: the educational format (no sales pitch — purely informational for buyers), your role (explaining the mortgage process, down payment programs, pre-approval steps) and theirs (explaining the buying process, how offers work, what to expect at inspection), the suggested format (in-person or virtual, 60-90 minutes), and the mutual benefit (building trust with buyer leads before they are under contract). Tone: collaborative, enthusiastic. Length: 200-250 words. ``` For more AI-powered scripts tailored to real estate professionals, see role prompts for realtors and AI for real estate.

Online reviews on Google and Zillow drive a measurable share of inbound mortgage leads. The problem is most loan officers ask for reviews too late, too generically, or not at all. ChatGPT can generate review-request messages tailored to the timing (right after closing, while the emotion is fresh) and the channel (text, email, or LinkedIn message).

Prompt — Post-closing review request email: ``` You are a mortgage loan officer. Write a short, genuine email asking a borrower to leave a Google review following the successful closing of their home purchase. The borrower's name is [NAME]. Reference the key milestone they just hit (closing day, getting keys). Make the ask feel personal, not automated. Include a direct sentence like: "If you have 2 minutes, a Google review would mean the world to me — it helps other families find someone they can trust." Provide a placeholder [YOUR GOOGLE REVIEW LINK] for the link. Do not offer any incentive for leaving a review (this violates review platform policies). Length: 80-120 words. ``` Send this within 24-48 hours of closing while the borrower is still in the glow of getting keys. The response rate is dramatically higher than a review request sent weeks later.

Prompt — Review request follow-up (for non-responders): ``` You are a mortgage loan officer. Write a brief, non-pushy follow-up message (suitable for text or email) to a past borrower who did not respond to an earlier review request. Acknowledge that they are probably busy settling into their new home. Make the ask light: "If you ever have a spare 90 seconds, a quick review helps more than you know." Include a placeholder for the review link. Tone: warm, zero pressure. Under 60 words. ```

Educational content — blog posts, email sequences, social media captions — positions a mortgage broker as a trusted resource before a buyer is ready to apply. This content is lower-risk from a compliance standpoint because it is general and educational rather than transactional, but it still needs a human review to ensure nothing constitutes a specific offer or advertisement under Regulation Z.

Prompt — "What is a debt-to-income ratio?" explainer article: ``` You are a mortgage loan officer writing a short educational article for first-time homebuyers on your website. Explain what a debt-to-income (DTI) ratio is, how it is calculated (using a simple example with round numbers, not a borrower's real figures), why lenders care about it, and what a borrower can do to improve their DTI before applying. Use plain English throughout. Do not reference specific DTI thresholds as if they are universal rules — note that thresholds vary by loan program and lender. Include a closing sentence encouraging the reader to speak with a licensed loan officer for guidance specific to their situation. Target length: 350-450 words. ```

Prompt — Down payment program social media carousel (5 slides): ``` You are a mortgage broker creating educational social media content for first-time homebuyers. Write copy for a 5-slide carousel post explaining common down payment assistance programs. Slide 1: hook — "You may not need 20% down to buy a home." Slides 2-4: cover FHA (minimum down payment described conceptually), state/local housing finance agency programs (describe the concept of forgivable second liens without naming specific programs that may vary by state), and conventional 97 options. Slide 5: call to action — speak with a licensed loan officer to find programs available in your area. Keep each slide to 2-3 sentences. Do not quote specific assistance amounts — those vary by program and change frequently. ```

Prompt — Email nurture sequence opener for leads who aren't ready to buy yet: ``` You are a mortgage broker writing the first email in a nurture sequence for a lead who inquired about home buying but said they are 6-12 months away from being ready. Write a warm, educational email that: acknowledges their timeline without pressure, offers one actionable tip they can implement right now to improve their mortgage readiness (e.g., checking their credit report — reference AnnualCreditReport.com as the free official source), and ends with an invitation to reach out when they have questions. Do not mention specific rates or loan programs. Tone: helpful, zero sales pressure. Length: 150-200 words. ```

Objection handling is one of the most underprepared skills in mortgage sales. When a borrower says "my bank can probably beat your rate" or "I want to wait until rates drop," most loan officers wing it. ChatGPT can help you build a library of thoughtful, honest responses that you rehearse and personalize — not scripts you read verbatim.

Prompt — Objection: "I'll just go with my bank": ``` You are a mortgage broker coaching a junior loan officer. Write a calm, honest, non-defensive response to the objection: "I think I'll just go with my bank — it's easier." The response should: acknowledge the convenience appeal of a bank relationship, explain the structural difference between a bank (one set of products) and a broker (access to multiple lenders), and offer a specific next step ("Let me run your scenario through a few lenders and show you the comparison — it takes 20 minutes and you're under no obligation"). Tone: confident, not pushy. Under 150 words. ```

Prompt — Objection: "I'm waiting for rates to drop": ``` You are a mortgage loan officer. A buyer says: "We've decided to wait. Rates might come down in the next few months." Write a thoughtful, honest response that: respects their autonomy, briefly explains the uncertainty of rate timing without making a rate prediction (you cannot know where rates are going, and neither can anyone else), and introduces the concept of buying now and refinancing if rates drop significantly. Do not predict or imply any specific rate movement. End with: "Would it make sense to at least get pre-approved so you're ready when the time comes?" Tone: consultative, no pressure. Under 200 words. ``` These scripts are internal training tools and personal rehearsal aids. They help you respond with consistency and calm when objections come up — which they always do.

Mortgage brokers receive a constant stream of agency guideline updates from Fannie Mae, Freddie Mac, FHA, and their wholesale lender partners. Reading and summarizing these bulletins is tedious and time-consuming. ChatGPT can compress a long guideline bulletin into a concise internal summary note in under a minute — but you must paste the original text into the prompt rather than asking the model to recall it from memory, since models can hallucinate guidelines they haven't actually seen.

Prompt — Guideline bulletin summarizer: ``` You are a mortgage underwriting assistant. Below is the text of a lender guideline bulletin. Summarize it for a team of loan officers in plain English. Cover: what changed, what the effective date is, which loan programs are affected, and what action the loan officer needs to take (if any). Flag anything that affects borrower eligibility or required documentation in bold. Keep the summary under 200 words. Do not add information that is not present in the bulletin text — only summarize what is written. [PASTE BULLETIN TEXT HERE] ``` This prompt pattern is important: the "do not add information not present in the text" instruction significantly reduces hallucination risk when the model is working from source material you have provided. Never use AI to recall or reproduce guideline text from its training data — agency guidelines change, and the model's training cutoff may predate recent updates. Always work from the official published source.

Prompt — Internal scenario notes for a file review: ``` You are a senior mortgage loan officer. Write internal scenario notes for a file review meeting. The borrower profile is: [DESCRIBE: e.g., self-employed 2 years, primary income from Schedule C, recent job gap, buying a primary residence in a non-recourse state]. Summarize the key risk factors a processor or underwriter should flag, the documentation that will likely be scrutinized, and the loan programs most likely to fit this profile. Format as bullet points under three headings: Risk Factors, Documentation Watch List, Program Considerations. This is an internal working document, not borrower-facing. ``` For a broader look at how AI can support financial workflows, see AI for finance teams and prompt engineering for finance teams.

These prompts give you significant leverage, but it is worth being explicit about where the boundary is. ChatGPT and similar models must not be used to: generate official loan disclosures (Loan Estimate, Closing Disclosure, or any Regulation Z disclosure), quote APR figures or specific interest rates as if they are offered terms, make or assist in making credit decisions (approving, denying, or conditioning a loan on any protected characteristic under ECOA), produce advertisements that trigger Regulation Z disclosure requirements without including the required disclosures, or create compliance documentation that a licensed compliance officer must prepare and sign.

The framework that keeps you safe is: AI drafts educational and internal content; humans approve everything borrower-facing; compliance reviews anything with regulatory implications. RESPA Section 8 prohibitions on kickbacks and fee-splitting also apply to co-marketing arrangements — if you use AI to draft a co-marketing proposal with a realtor, the actual arrangement still must comply with RESPA. The AI draft does not change the underlying legal requirement.

The CFPB publishes plain-language guidance on TILA, RESPA, and ECOA requirements at consumerfinance.gov. For teams building automated prompt workflows — for example, a CRM integration that auto-drafts borrower update emails — the OpenAI prompt engineering guide at platform.openai.com/docs/guides/prompt-engineering covers how to constrain model outputs systematically. Use both. Want to understand the cost of running these prompts at scale? Check our AI Prompt Cost Calculator to estimate your monthly usage cost before building any automation.