A key element of the proposal is the push for a single federal normal, with officers warning {that a} patchwork of state-level laws “would undermine American innovation and our capability to steer within the world AI race.”
“The Administration seems ahead to working with Congress within the coming months to show this framework into laws that the President can signal,” the announcement said.
Diane Yu, CEO of mortgage expertise platform Tidalwave, instructed HousingWire {that a} federal normal would speed up AI adoption within the business.
“We have now lenders in all 50 states utilizing our point-of-sale platform at this time. Which means each new state-level AI regulation is one thing our compliance and engineering groups have to judge, interpret and construct for, as an alternative of spending that point making the product higher for debtors and mortgage officers,” Yu stated. “A single federal normal would allow us to make investments that power into the product itself.”
Yu stated she’s had conversations with lending executives who’re able to undertake AI however are caught in authorized overview as a result of their counsel can’t give them clear solutions for what’s required on a state-by-state foundation.
“Lenders already comply with one nationwide set of underwriting pointers by Fannie Mae and Freddie Mac. Including a second, fragmented layer of AI-specific state guidelines on prime of that doesn’t create readability. It creates paralysis,” she stated, including that “the route [of the framework] is correct” and that the mortgage business is a federally regulated business at its core.
“Lenders already adjust to ECOA, Truthful Housing Act, TILA, and RESPA. The GSEs set underwriting requirements nationally by Fannie Mae and Freddie Mac. And the federal guardrails aren’t theoretical. Freddie Mac’s Bulletin 2025-16, which went into impact March 3, requires each vendor/servicer to determine a complete governance framework for AI and machine studying techniques,” Yu stated.
“Which means steady monitoring, formal bias testing, alignment with NIST and ISO cybersecurity requirements, senior administration accountability, segregation of duties between AI improvement and threat oversight, and impartial audits. That’s not gentle contact. That’s rigorous, mortgage-specific AI oversight, and it’s already taking place on the federal degree.”
Whereas Yu is optimistic concerning the ethos of the framework, she stated she wished the framework addressed mortgage-specific AI use instances.
“One factor that’s lacking from many of the AI coverage dialog is the excellence between corporations which are really deploying AI in manufacturing and corporations which are simply speaking about it. In mortgage, that hole is vast. We’ve been in manufacturing with nationwide lenders for over a 12 months. That’s a distinct dialog [from] an organization displaying a demo at a convention,” Yu stated.
“I’d additionally wish to see the framework deal with one thing particular to mortgage: the business already loses roughly $600 per mortgage on origination prices, and shutting a mortgage nonetheless takes 43 days on common,” she added. “If the aim is to guard shoppers, quicker approvals, decrease prices, and fewer errors are [key to] shopper safety. The appropriate regulatory framework ought to make it simpler for lenders to undertake expertise that delivers these outcomes, not tougher.”
