How State vs. Federal AI Policy Plays Out In 2025

A key thing we learned spending time with Sagent customers at our Ignite user conference this spring was that all mortgage servicers must know how AI regulation – and particularly state vs. Federal AI regulation – will play out in 2025. So, here’s the latest as of early-summer.  

The Evolution of AI Executive Orders – 2023 To Present 

In October of 2023, the White House issued Executive Order 14110 emphasizing infrastructure and standards by establishing the U.S. AI Safety Institute within the National Institute of Standards and Technology (NIST). This directive aimed to position America as a global leader in AI governance, safety, and security. 

In January 2025, Executive Order 14110 was replaced by Executive Order 14179, which aims to streamline regulation to foster American innovation and investment. This approach also framed American leadership in AI as a matter of national security. 

However, because Executive Orders are primarily directive and lack the force of Federal legislation, states have taken the initiative to implement their own AI-related policies—which adds an emerging patchwork of new state-level measures for mortgage lenders and servicers. 

State-Level AI Policies Proliferate  

Colorado was first to pass a comprehensive AI bill May 17, 2024, addressing developers and deployers of “high risk” AI systems that make, or are a substantial factor in making, a consequential decision pertaining to education, employment, financial or lending services, essential government services, healthcare, housing, insurance or legal services.   

This law was broad and deep, with sweeping definitions and a battery of risk assessment, testing, affirmative disclosure and other requirements. It was also accompanied by a signing letter from the Governor’s Office, acknowledging that while a uniform federal approach would be the most efficient, Colorado would be protecting its own citizens in the meantime. However, the Governor also provided a list of requested amendments to clarify the scope and refine the requirements of the law prior to its effective date of February 1, 2026.   

California, Virginia, Utah, Texas, and Montana, to name a few, have either passed or introduced similar legislation addressing AI.  

And others, like Maryland and Kentucky, have established commissions to develop proposed policies and standards.  

Overall, the National Council of State Legislatures is tracking more than 40 state measures across the country relating to AI. 

State vs. Federal AI Policy Tension Building 

As state-level AI legislation has grown, on May 22, 2025, the U.S. House of Representatives voted to pass a 10-year moratorium on the enforcement of state laws governing AI.  

In response, a bipartisan group of 260 lawmakers from 50 states submitted a letter calling upon Congress to reject the ban on state enactment and enforcement of AI law and regulation. They claim the measure will impede the states’ ability to protect their residents from AI-related harm or thwart local technology industries and nimble state response to emerging issues in AI use or innovation.  

The moratorium has sparked significant continuing debate and has recently come before the Senate.  The procedural hurdle of the Byrd Rule (which prohibits non-budget items from inclusion in budget-related legislation) was cleared, and Senate discussion is anticipated to cover the gamut of economic and technological and pluralism concerns. 

On June 4, 2025, the White House issued a news release setting forth the intention to change the name of the U.S. AI Safety Institute to the Center for AI Standards and Innovation (CAISI). This rebranded entity will: 

• Work with NIST organizations to develop guidelines and best practices to measure and improve the security of AI systems 

• Work with the NIST Information Technology Laboratory and other NIST organizations to assist industry to develop voluntary standards 

• Establish voluntary agreements with private sector AI developers and assess AI capabilities that may pose risks to national security (cybersecurity, biosecurity, and chemical weapons) 

• Study the capabilities of U.S. and adversary AI systems, the adoption of foreign AI systems, and assess international AI competition 

• Assessments potential security vulnerabilities associated with AI systems, including the possibility of backdoors and other covert, malicious behavior 

• Coordinate with other federal agencies and entities, including the Department of Defense, the Department of Energy, the Department of Homeland Security, the Office of Science and Technology Policy, and the Intelligence Community, to develop evaluation methodology and conduct evaluations and assessments 

• Represent U.S. interests internationally in the regulation of American technologies by foreign governments and collaborate with the NIST Information Technology Laboratory to maintain US influence over international AI standards 

As July kicks off, the Senate voted to remove the 10-year moratorium from the tax and spending package, which is the latest in this state-federal AI regulation debate. It will continue to play out in real time, and we’ll provide a new briefing as the policy topics above are worked out between Washington and the states.  

Sagent and our teams are all over these AI policy topics as we build the future of mortgage servicing, so if you have questions about AI policy impacts for mortgage servicers, please submit the form below, and we’ll reply right away.