A compliance-focused webinar held on Thursday, April 2, highlighted increasing risks associated with the rapid adoption of artificial intelligence (AI) across the mortgage industry. Speakers warned that many lenders and loan originators are implementing AI faster than their compliance frameworks can support.
The session, titled "Maximum Compliance — A MaxClass Series: The April Rules to not be a Compliance Fool: This Month’s Sticky Situations," featured industry educators LaDonna Lockard and Jeana Lanktree. They detailed AI's pervasive presence in mortgage workflows, from marketing and prequalification to income calculations.
“It is literally everywhere,” Lockard said. “It’s now in our marketing copy. It’s in our pre-qual conversations.”
Lockard added that many organizations are only now recognizing their unpreparedness for the compliance implications. “It feels like even though we knew this was coming, we’re all kind of being slapped upside the head right now … we didn’t put everything in place that we needed to be prepared for this.”
Two Camps Emerge As AI Use Accelerates
Lanktree described a growing divergence in how lenders approach AI adoption. “We have two camps right now,” she said. “We have the camp that is fully integrating … and then we have the other camp that’s resisting.”
She noted that a "happy medium" between these extremes remains underserved, particularly concerning training. Operational pressures, however, are driving AI adoption. Loan originators and processors increasingly use AI tools to meet turn times, calculate income, and manage complex loan scenarios, including bank statement and Debt Service Coverage Ratio (DSCR) loans.
“People are trying to meet turn times … and now they’re integrating that into AI,” Lanktree said.
Data Exposure And Disclosure Risks Intensify
A primary concern discussed during the session was the handling of borrower data. Lanktree stated that many originators unknowingly expose sensitive information when using AI tools.
“Are you redacting all of the client’s personal information?” she asked, noting the "shock moment" when most participants realize they are not.
Lockard warned that uploading full loan files into open AI systems creates significant compliance exposure. “Your information is out there. It can be searched, it can be disclosed, it can be hacked,” she said.
Beyond data privacy, speakers pointed to risks from AI-generated outputs, including inaccurate disclosures, misleading product comparisons, and potential fair lending violations. “What happens then? Who’s on the hook?” Lockard asked.
Liability Remains Squarely On Lenders And Originators
Despite the increasing use of AI, responsibility for errors has not shifted. “If you quote something wrong … you’re on the hook for that,” Lanktree said.
This liability applies whether an issue stems from human error or AI-generated output, especially for smaller broker shops without dedicated compliance or technology teams. Ensuring accuracy across multiple AI tools, often used simultaneously, can be challenging.
“You have so many options right now … ensuring accuracy becomes nearly impossible,” Lanktree said.
Bias And Misinformation Remain Unresolved Challenges
Speakers also highlighted the risk of bias within AI systems, noting that outputs can reflect underlying data and user prompts. “It picks up on our personality … what you put into that is going to directly correlate to what your response is,” Lanktree said.
AI-generated advice can also introduce misinformation into transactions, complicating deals and undermining borrower confidence.
Compliance Gaps Extend Beyond AI
While AI dominated the discussion, the session also addressed broader compliance pitfalls already impacting the industry. Lockard cited a recent case where a loan originator failed to receive course credit after not signing required documentation, ultimately triggering licensing and potential SAFE Act issues.
“These are those little compliance traps that really can get people in a whole lot of trouble down the line,” she said.
Lanktree pointed to another growing issue: unlicensed activity by loan partners taking on expanded roles. “That’s where things get sticky … you could be in a world of a mess,” she said.
Human Connection Seen As Long-Term Differentiator
Despite the rapid expansion of AI, both speakers emphasized that human interaction will remain critical and may become a competitive advantage. “The need for real contact … is going to become more critical than ever,” Lanktree said.
As AI-generated interactions become more difficult to distinguish from real ones, borrowers may increasingly seek direct, human engagement. Lockard added that originators who establish expertise and trust will be better positioned as technology continues to evolve.
Training And Guardrails Remain Key Priorities
The webinar concluded with a call for greater industry-wide focus on compliance training, data protection, and internal controls. The rapid pace of AI adoption has created a gap between capability and oversight, which lenders and originators must close quickly.
“There aren’t necessarily guardrails in place,” Lockard said. “That’s where we need to come into the picture and actually implement those things.”
