Skip to content

Nearly Half of Americans Are Comfortable With AI Approving or Denying Their Mortgage

Updated: Sept 16 2026 • 6 min read

Smiling couple sitting on a couch using a laptop and smartphone together.

Key Findings

  • Nearly half of consumers (49.2%) are comfortable with AI approving or denying a mortgage application. Comfort is even higher for other parts of the mortgage process, including determining qualification (64.7%) and recommending how much to borrow (67.3%).
  • Comfort with AI doesn't necessarily mean consumers want it making every decision on its own. 38.0% want AI to make recommendations with a human making the final decision, while 18.8% are comfortable with AI making some decisions independently with human review of major decisions.
  • Consumers who have already used AI for mortgage or loan questions are especially open to AI taking on a bigger role. Among this group, 72.1% are comfortable with AI approving or denying a mortgage application, compared with 41.1% of those who haven't used AI for mortgage or loan questions.
  • When a problem could put closing at risk, consumers still want a person involved. 54.4% would turn to a human loan officer and another 28.4% would want a human loan officer and AI working together.

AI is already becoming part of how people research financial decisions. But when it comes to getting a mortgage, consumers appear willing to let it go much further than answering questions or comparing rates.

To understand how far consumers are comfortable letting AI into the mortgage process, Lower surveyed 1,000 U.S. consumers about how they feel about AI handling everything from comparing rates to approving or denying a mortgage application, and where they still want a person involved.

According to the survey, nearly half of consumers (49.2%) surveyed by Lower say they're comfortable with AI approving or denying a mortgage application. And comfort extends across several other consequential parts of the process. Majorities are comfortable with AI recommending how much they should borrow (67.3%), determining whether they qualify for a mortgage (64.7%), and even reviewing their financial documents (54.3%).

At the same time, consumers aren't ready to hand over the entire mortgage process. When asked what role AI should play overall, the most common response was that AI can make recommendations, but a human should make the final decision.

AI Is Moving Deeper Into the Mortgage Process

Lower asked consumers how comfortable they'd be with AI handling different tasks across the mortgage process, from answering general questions to approving or denying an application.

Mortgage task

Comfortable with AI

Comparing mortgage rates and loan options

76.8%

Answering general mortgage questions

74.9%

Recommending a specific type of mortgage

71.8%

Estimating how much home you can afford

70.7%

Recommending how much you should borrow

67.3%

Determining whether you qualify for a mortgage

64.7%

Reviewing financial documents

54.3%

Approving or denying a mortgage application

49.2%

Comfort is highest for uses like comparing rates and answering mortgage questions, but it also remains relatively high as AI moves further into the process.

More than seven in 10 consumers are comfortable with AI recommending a mortgage type, and roughly two-thirds are comfortable with it recommending how much they should borrow or determining whether they qualify. Comfort remains above 50% even when AI is handling sensitive financial information, with 54.3% comfortable with AI reviewing their financial documents.

Approving or denying a mortgage application is the only task where comfort falls below 50%. But at 49.2%, it's also the clearest sign that consumers' comfort with mortgage AI extends well beyond research and comparison.

Comfort With AI is Higher Among Mortgage AI Users

Comfort with AI approving or denying a mortgage application is not evenly distributed.

Among consumers who have already used AI for a mortgage or loan question, 72.1% are comfortable with AI approving or denying a mortgage application, compared with 41.1% of those who haven't used AI for that purpose.

Mortgage experience shows a similar pattern. Among current mortgage holders, 58.4% are comfortable with AI approving or denying a mortgage application, compared with 33.0% of consumers who have never had a mortgage and don't plan to get one.

That higher comfort may come from having already seen what AI can do. Early in the process, AI can help answer an initial question like whether homeownership is even possible before a consumer speaks with a loan officer.

“Once AI has answered the question and saved someone real time, they’ll trust it with more,” said Gino Fronti, VP of Product for LOAI at Lower. “People have seen it work.”

Comfort With AI Isn't the Same as Handing Over the Mortgage

Being comfortable with AI approving or denying a mortgage application doesn't necessarily mean consumers want AI making every mortgage decision on its own.

When asked what role AI should play overall, 38.0% say AI can make recommendations, but a human should make the final decision. Another 18.8% are comfortable with AI making some decisions independently as long as major decisions still receive human review.

Far fewer want AI operating mostly on its own. Just 5.2% are comfortable with AI making most mortgage decisions independently, while 4.3% would let AI handle the entire mortgage process without human involvement.

For Fronti, that distinction reflects what he's seeing from consumers.

“People like to start the mortgage process online, from getting their questions answered to getting pre-approved. But at some point they want an advisor. They want somebody in an advisory role before they make the final decision,” Fronti said. “This is the largest investment of most people's lives. Many people don’t want to make that decision entirely inside a browser window.”

That distinction helps put the 49.2% approval-or-denial finding into context. Consumers can be comfortable with AI taking on consequential parts of the mortgage process without wanting to remove people from the process entirely.

When Something Goes Wrong, Consumers Still Want a Person

Consumers' preference for human involvement becomes clearest when there's a problem.

When asked who they'd most want to help resolve an unexpected issue that could delay or prevent closing, 54.4% chose a human loan officer. Another 28.4% chose a human loan officer and AI working together. Just 7.0% chose an AI-powered mortgage assistant alone.

Combined, 82.8% selected an option that includes a human loan officer when a problem could put closing at risk.

“When something threatens closing, consumers overwhelmingly want a human involved. Some want a loan officer directly, while others want a loan officer and AI working together. That tells us technology can play a bigger role without replacing the human relationship,” Fronti said.

Access matters beyond problems, too. When asked what would make them more comfortable with AI during the mortgage process, 49.1% of consumers selected having access to a human whenever they want one, the most commonly selected response. Another 38.1% selected having a human loan officer review important AI recommendations or decisions.

Even with relatively high comfort across many mortgage tasks, the findings show how important human involvement remains when something could put closing at risk.

What This Means for Homebuyers

AI is likely to show up in more parts of the mortgage process, whether it's helping compare loan options, estimating affordability, or reviewing information along the way. For homebuyers, that could mean more ways to get information and move through parts of the process faster.

But comfort with AI doesn't have to mean choosing between technology and a loan officer. The survey shows consumers are open to both. Nearly half are comfortable with AI making an approval or denial decision, while access to a human remains important when they want help or something goes wrong.

For loan officers, that could mean spending less time on routine parts of the process and more time advising borrowers.

“Technology is taking the paperwork off our plate, not the relationship,” Fronti said. “It frees a loan officer up to do the part that actually takes judgment, which is advising.”

For buyers, the question may be less about whether AI belongs in the mortgage process and more about where it's useful to them, where they want another set of eyes, and when they'd rather talk to a person.

Methodology

Lower surveyed 1,000 U.S. consumers through Pollfish on September 9, 2026, about their comfort with AI in the mortgage process, including specific mortgage tasks, AI's role in decision-making, and when they would want a human involved. Percentages reflect Pollfish's stratified, weighted results. For questions that allowed respondents to select more than one answer, percentages represent the share who selected each response and may not sum to 100%.

This analysis is for informational purposes only. Contact a Lower loan expert for a personalized rate and payment estimate.

Explore your home loan options.

Mortgage Resources

Clear
Selection