AI for Mortgage Underwriter
A single self-employed borrower file can consume 2–3 hours of your morning, and declination notices that require ECOA-precise language take another 20–30 minutes each — on top of 30–60 minutes a day explaining the same conditions to processors via email. These guides show you how to cut through the writing-heavy parts of the job: from income narratives to condition explanations to adverse action letters, with the precision the role demands and less time staring at a blank document.
Ready to try? Start with a prompt →
Updated 14 days ago
New to AI?
The Big Four AI Assistants
ChatGPT, Claude, Gemini, and Grok do roughly the same thing. Pick one and start.
Four Levels of AI Skill
From your first prompt to building automated workflows. Where are you now?
How to Keep Up with AI
The landscape changes fast. A low-effort system to stay informed without drowning.
Try right now
Copy a prompt, paste into ChatGPT, Claude, or Gemini
Works with any free AI chatbot — no signup needed
Draft an Adverse Action Notice
An ECOA/FCRA-compliant adverse action notice with proper legal language citing specific denial reasons, ready to fill in borrower details and send within regulatory timelines.
Draft an ECOA/FCRA compliant mortgage adverse action notice. Denial reasons: [list specific adverse factors, e.g., "debt-to-income ratio exceeds maximum guideline" and "insufficient cash reserves"]. Loan type: [conventional/FHA/VA]. Leave blanks for applicant name, date, and application number.
Tip: List the denial reasons as precisely as the regulation requires — "excessive obligations in relation to income" is the ECOA-standard phrasing for high DTI. Ask the AI for the correct regulatory phrasing if you're unsure: "What is the ECOA-standard adverse action reason for [scenario]?"
Summarize an Agency Guideline Bulletin
A concise 1-page summary of a Fannie Mae, Freddie Mac, FHA, or VA guideline bulletin — key changes, effective dates, and action items — so you and your team can digest it in 5 minutes instead of 2 ...
Summarize the following mortgage agency guideline bulletin. Extract: 1) key policy changes, 2) effective dates for each change, 3) action items for underwriters and processors, and 4) any file types or scenarios most impacted. Format as a brief 1-page summary. Bulletin text: [paste bulletin text here]
Tip: After summarizing, ask "What are the top 3 things an underwriting team needs to do differently starting on the effective date?" — this turns a summary into an action plan you can share directly with your team.
Generate a Compensating Factors Checklist for a Borderline File
A comprehensive checklist of all compensating factors that could support approval of a borderline mortgage file — organized by risk area — so you don't miss any factor that could make the differenc...
Generate a comprehensive compensating factors checklist for a borderline [conventional/FHA/VA] mortgage file with [describe the specific risk — e.g., "high DTI at 47%" or "thin credit file with only 3 tradelines"]. List every compensating factor that could support approval, organized by category: credit, assets, income stability, LTV, payment shock, employment. Include the Fannie Mae or FHA standard for each where applicable.
Tip: Run this before you render a decision on any borderline file — sometimes you'll identify a compensating factor sitting in the file that you hadn't weighted appropriately. It's also useful for the exception request narrative (see that prompt guide) — use this checklist output as the raw material.
Write a Clear Condition Letter for Processors
A professional, detailed condition letter that tells processors exactly what documentation is needed, what it should show, and why — reducing back-and-forth and re-submissions.
Write a mortgage underwriting condition letter requesting [specific documentation, e.g., "12 months business bank statements for a self-employed borrower"]. Explain: what the document must show, acceptable format, and why it's required for underwriting. Write clearly enough for a processor who is new to this condition type.
Tip: If you're regularly getting the same type of incomplete document back, add "and note the 3 most common mistakes that make this document unacceptable" to your prompt — you'll get a condition letter that pre-empts the typical errors.
Use AI in your tools
AI features built into tools you already have
AI features already built into your existing tools
Use Excel Copilot to Build and Verify Income Calculation Worksheets
Excel Copilot helps you build income calculation worksheets, write formulas using plain English, and check your existing calculations for errors — reducing the risk of income miscalculations that c...
Use Guideline Buddy to Search Agency Guidelines in Plain Language
Guideline Buddy is a purpose-built AI tool for mortgage professionals that lets you search Fannie Mae, Freddie Mac, FHA, VA, and USDA guidelines using plain-language questions — instead of navigati...
Use Outlook Copilot to Summarize Loan File Email Threads
Outlook Copilot reads the entire email thread for a loan file and produces a bullet-point summary of what happened — what conditions were issued, what was received, what's still pending, and who co...
Use Word's AI to Draft Underwriting Letters and Narratives
Word's built-in Copilot AI drafts adverse action notices, exception request narratives, and underwriting summary letters from a brief description — eliminating the blank-page problem for complex wr...
Set up an AI assistant
Step-by-step guides for dedicated AI tools
10–30 minute setup, then ongoing time savings
Use ChatGPT to Speed Up Complex Income Analysis
By the end of this guide, you'll have a system for using ChatGPT to accelerate complex income analysis — structuring your analysis plan, double-checking your methodology, summarizing findings for t...
Build a Reusable Adverse Action Notice Library with Claude
By the end of this guide, you'll have a complete library of ECOA/FCRA compliant adverse action notice templates — one for each major denial scenario — saved and ready to use.
Configure ChatGPT as Your Mortgage Underwriting Assistant
By the end of this guide, you'll have ChatGPT configured with your lender's specific context — loan types, guideline sources, common file types, and preferred writing style — so every session produ...
Build a Claude Project Guideline Research Assistant
By the end of this guide, you'll have a Claude Project loaded with the Fannie Mae Selling Guide, Freddie Mac Single-Family Guide, and your lender's overlay documents — so you can ask guideline ques...
Go further
Advanced workflows, automation, and custom AI setups
For when you’re ready to connect tools and automate
Automate Pipeline SLA Alerts: Connect Your LOS to AI-Drafted Notifications
An automated system that watches your pipeline for files approaching SLA deadlines and sends AI-drafted alert messages to processors and loan officers — without you manually tracking or writing eac...
Build a Custom GPT That Answers Agency Guideline Questions for Your Whole Team
A shared Custom GPT loaded with Fannie Mae, Freddie Mac, FHA, and your lender's overlay documents — configured specifically for mortgage underwriting — that every underwriter on your team can acces...
Recommended Tools
3Ranked by relevance for mortgage underwriter
Claude
Draft Adverse Action / Declination Notices, Draft Condition Letters and Processor Instructions + 5 more
ChatGPT
Write Exception Request Narratives, Draft Loan File Notes and Underwriting Rationale + 2 more
Microsoft Outlook
Use Outlook Copilot to Summarize Long Email Threads
This guide is refreshed as tools evolve. Bookmark it.
Last updated 14 days ago