Question of the Week: How Does Our Industry Use Artificial Intelligence?

Artificial intelligence (AI) is broadly defined as computer systems that can perform tasks that normally require human intelligence.

While “AI” has become a buzzword in recent years, the mortgage industry has quietly used AI‑type systems for decades—long before the label existed.

Below are the major areas where AI shapes our world today.

Credit

Credit scoring has existed as long as banking itself. Even before numeric scores, lenders evaluated character, capacity, collateral, and credit history to determine whether someone was a good risk.

1956 — The Birth of Modern Credit Scoring

In 1956, Bill Fair and Earl Isaac founded Fair, Isaac and Company (FICO). (Fun fact: many people assume “Fair” refers to fairness in scoring—it’s just Bill Fair’s last name.)

Their innovation was applying data analytics to estimate credit risk, assigning numeric values to borrower behaviors. Lenders subscribed to their scoring system, which brought consistency to an otherwise subjective process.

Late 1950s–1980s — Computerization

Using punch cards and early computers (cue one of your favorite movie/acting duos), lenders began automating credit scoring. As credit bureaus grew and consolidated, consistent scoring became essential.

1989 — The First Standardized FICO Score

FICO released a score based on data from all three major credit bureaus. For the first time, lenders across the country could use the same standardized measure of credit risk.

AI’s Role in Credit Today

Over time, scoring models evolved by analyzing millions of accounts, identifying patterns in repayment behavior, and predicting future risk.

This is where AI became foundational—learning from past and current behavior to assign risk with far more accuracy than manual review ever could.

Underwriting

When I started in the industry in 1987, underwriting was a very different world.

• We relied on couriers and the U.S. Postal Service.
• Fax machines weren’t standard.
• Underwriters required original documents—tax returns, bank statements, everything.
• A loan decision typically took two weeks.
• Even credit reports could take up to two weeks, and processors eagerly awaited the daily delivery of reports.

The Fax Machine Revolution

The fax machine was a game‑changer. Suddenly we could send documents to underwriters in minutes. Direct communication improved, and when we finally got a printer that could print credit reports in‑office, it felt like magic.

Then came the internet and email, which transformed communication entirely.

1995 — The Leap to Automated Underwriting

In 1995, our industry took a massive step forward. Fannie Mae and Freddie Mac introduced Automated Underwriting Systems (AUS)—statistically based models that could analyze a loan file and issue a decision.

At first, underwriting departments entered the data manually. Even that cut turn times dramatically.

1999 — AUS Goes Online

In 1999, Fannie and Freddie made AUS available via the internet. Brokers, lenders, and wholesalers could input borrower information and receive decisions in minutes.

Today

A single conversation with a borrower can produce:

• a credit report
• a complete application
• an AUS decision

…all during the phone call.

What once took an underwriter hours to review, calculate, and decide, AUS now does in under a minute.

Rates

AI does not set mortgage rates. But it plays a major role in how lenders interpret the data that influences rates.

Mortgage rates are shaped by:

• inflation
• bond yields
• Federal Reserve policy
• labor markets
• investor demand
• risk
• consumer behavior
• global events

Even with all this information, rate setting ultimately comes down to educated forecasting—what lenders and investors believe the economy will do next.

Where AI Fits In

AI excels at:

• analyzing massive datasets
• monitoring markets in real time
• identifying patterns in economic releases
• comparing current conditions to historical movements

As markets shift throughout the day, AI helps lenders decide whether to raise, lower, or hold rates. In today’s volatile environment—energy prices, geopolitical conflict, instant news—we’re seeing more intraday rate changes as lenders try to stay ahead of the curve.

The Future of AI in the Mortgage Industry

AI performs tasks traditionally done by humans. If the technology continues to advance, parts of the mortgage process could become fully automated—and the industry could contract significantly.

Imagine this future:

Dennis and Leslie need a mortgage. They log onto a website, create an account, and complete the standard application. They click a button—their credit report is pulled. They click another—their application is uploaded to Fannie Mae or Freddie Mac. Seconds later, they receive a loan decision.

Next, they click a button to put their application on the market. Their account begins filling with lender offers, each with rates and terms. They choose a lender, lock their loan, and upload income and asset documents.

The lender’s AI system:

• analyzes the documents
• issues disclosures
• determines whether an appraisal is needed
• clears conditions
• prepares the file for funding

…all in a matter of days.

Will humans still matter?

Absolutely—but fewer of us.

There will always be borrowers who need guidance, problem‑solving, and human expertise. But many will choose the fully automated path, especially as lenders continue offering direct‑to‑consumer digital mortgage platforms.

Sorry to my fellow mortgage professionals—but we all know this future is coming.