One of the country’s largest mortgage lenders is turning to artificial intelligence to answer the phone. New American Funding said it has partnered with Kastle, an enterprise AI platform built for consumer lending, to deploy voice AI across both its mortgage origination and servicing operations, according to an announcement Wednesday.
The move reflects how quickly AI is moving from the back office to direct customer contact in the mortgage industry, where high call volumes and heavy compliance requirements have long made phone support expensive and hard to scale.
What the technology does
Kastle’s system handles end-to-end customer phone interactions throughout the life of a loan, the companies said, reducing hold times and providing 24/7 self-service support. Its pre-built agents use what Kastle calls Safe Execution Procedures to complete multi-step mortgage transactions consistently while embedding regulatory requirements at the point of execution β an approach aimed at the industry’s central worry about automating regulated work.
By routing routine calls to the AI, New American Funding said its operations teams can focus on more complex situations that require human judgment. The lender did not disclose call-volume figures or other metrics for the rollout.
“Kastle AI enables us to serve each customer with greater speed and consistency throughout the life of their loan,” said Christy Bunce, president of New American Funding.
A large lender’s AI bet
Founded in 2003, New American Funding employs more than 5,300 people across over 300 locations, services a portfolio of more than 281,000 customers and holds roughly $73 billion in servicing value β scale that makes customer-service efficiency a meaningful cost lever.
“New American Funding understands that winners will be lenders who embrace AI without compromising compliance or control,” said Rishi Choudhary, co-founder and chief executive of Kastle.
Automation spreads across real estate
The partnership is part of a broader wave of AI adoption across housing and real estate, from brokerages rolling out AI operating platforms for agents to lenders automating underwriting and servicing workflows. The appeal is clear in an industry squeezed by high costs and thin margins, but the technology’s track record on accuracy remains a live question β testing of AI mortgage assistants has shown they can still get a meaningful share of consumer questions wrong, underscoring why compliance safeguards are central to any deployment that touches borrowers directly.
What it means: Handing customer calls to AI is a bigger step than automating internal paperwork, because it puts the technology directly in front of borrowers on regulated matters like payments, escrow and loan terms. For a servicer of New American Funding’s size, even modest gains in call handling translate into real savings and, potentially, faster response times for customers. The key test will be whether the AI can maintain accuracy and compliance at scale; lenders that get it right stand to lower costs materially, while missteps in a heavily regulated business carry outsized risk. Either way, the deployment is a marker of how central AI has become to the economics of mortgage lending.
