AI underwriting to human underwriting and this is what happened

Didier Malagies • September 24, 2025


 Speed & Efficiency


AI Underwriting:


Processes applications in seconds to minutes.


1.Can instantly pull data from multiple sources (credit reports, bank statements, income verification, property valuations, etc.).


Ideal for high-volume, standardized cases.


Human Underwriter:


Takes hours to days, depending on complexity.


Manually reviews documents, contacts third parties, and applies professional judgment.


Slower, especially for complex or edge cases.


2. Data Handling


AI:


Uses algorithms and machine learning to analyze massive datasets.


Can detect patterns humans might miss (e.g., spending behavior, alternative data like utility payments, even digital footprints in some markets).


Human:


Relies on traditional documentation (pay stubs, tax returns, appraisals).


Limited by human bandwidth—can’t process as much raw data at once.


3. Consistency & Bias


AI:


Decisions are consistent with its rules and training data.


However, if the data it’s trained on is biased, the system can replicate or even amplify those biases.


Human:


Brings subjective judgment. Can weigh special circumstances that don’t fit a neat rule.


Risk of inconsistency—two underwriters might interpret the same file differently.


May have unconscious bias, but also flexibility to override rigid criteria.


4. Risk Assessment


AI:


Excels at quantifiable risks (credit scores, loan-to-value ratios, historical claim data).


Weak at unstructured or nuanced factors (e.g., a borrower with an unusual income stream, or a claim with unclear circumstances).


Human:


Strong at contextual judgment—understanding unique borrower situations, exceptions, or “gray areas.”


Can pick up on red flags that an algorithm might miss (e.g., forged documents, conflicting information).


5. Regulation & Accountability


AI:


Regulators are still catching up. Requires transparency in decision-making (explainable AI).


Hard to appeal an AI decision if it can’t explain its reasoning clearly.


Human:


Provides a clear chain of accountability—borrower can request explanations or escalate.


Easier for compliance teams to audit decision-making.


6. Cost & Scalability


AI:


Scales cheaply—one system can process thousands of applications simultaneously.


Lower ongoing labor costs once implemented.


Human:


Labor-intensive, costs grow with volume.


Better suited for complex, high-value, or unusual cases rather than mass processing.


✅ Bottom line:


AI underwriting is best for speed, scale, and straightforward cases.


Human underwriters are best for nuanced judgment, exceptions, and handling edge cases.


Most modern institutions use a hybrid model: AI handles the bulk of simple files, while humans step in for complex or flagged cases.


tune in and learn https://www.ddamortgage.com/blog


didier malagies nmls#212566

dda mortgage nmls#324329




Ask a Mortgage Question

Use the form below and we will give your our expert answers!

203H Ask A Question

Start Your Loan with DDA today

Your local Mortgage Broker

Mortgage Broker Largo
See our Reviews

Looking for more details? Listen to our extended podcast! 

Check out our other helpful videos to learn more about credit and residential mortgages.

Will 6% interest rate create more activity for homebuyers

By Didier Malagies • November 5, 2025
This is a subtitle for your new post

My prediction for what is coming next

By Didier Malagies • November 3, 2025
Here are the main types of events that typically cause the 10-year yield to drop: Economic slowdown or recession signs Weak GDP, rising unemployment, or falling consumer spending make investors expect lower future interest rates. Example: A bad jobs report or slowing manufacturing data often pushes yields lower. Federal Reserve rate cuts (or expectations of cuts) If the Fed signals or actually cuts rates, long-term yields like the 10-year typically decline. Markets anticipate lower inflation and slower growth ahead. Financial market stress or geopolitical tension During crises (wars, banking issues, political instability), investors seek safety in Treasuries — pushing prices up and yields down. Lower inflation or deflation data When inflation slows more than expected, the “real” return on Treasuries looks more attractive, bringing yields down. Dovish Fed comments or data suggesting easing ahead Even before actual rate cuts, if the Fed hints it might ease policy, yields often fall in anticipation. tune in and learn https://www.ddamortgage.com/blog didier malagies nmls#212566 dda mortgage nmls#324329

Feds cutting rates what will happen?

By Didier Malagies • October 27, 2025
🏦 1. Fed Rate vs. Market Rates When the Federal Reserve cuts rates, it lowers the federal funds rate — the rate banks charge each other for overnight loans. That directly affects: Credit cards Auto loans Home equity lines of credit (HELOCs) These tend to move quickly with Fed changes. 🏠 2. Mortgage Rates Mortgage rates are not directly set by the Fed — they’re more closely tied to the 10-year Treasury yield, which moves based on investor expectations for: Future inflation Economic growth Fed policy in the future So, when the Fed signals a rate cut or actually cuts, Treasury yields often fall in anticipation, which can lead to lower mortgage rates — if investors believe inflation is under control and the economy is cooling. However: If markets think the Fed cut too early or inflation might return, yields can actually rise, keeping mortgage rates higher. So, mortgage rates don’t always fall right after a Fed cut. 📉 In short: Fed cuts → short-term rates (credit cards, HELOCs) usually fall fast. Mortgage rates → might fall if inflation expectations drop and bond yields decline — but not guaranteed. tune in and learn https://www.ddamortgage.com/blog didier malagies nmls#212566 dda mortgage nmls#324329 
Show More