Federal vs. State Tension in AI Regulation: What Healthcare Leaders Need to Know
- May 13
- 9 min read
By Dr. Ernest Wayde, PhD, MIS
As AI tools move deeper into clinical and operational workflows, many healthcare leaders are running into a growing problem as they try to adapt: the regulatory environment surrounding AI is getting more complex and unpredictable each month. While states are moving forward with their own AI laws and oversight frameworks, the federal government is pushing for a unified national approach. For healthcare organizations operating across multiple states, the result is a confusing and complicated and unsettled contested regulatory landscape that may already be affecting how organizations evaluate AI vendors, structure governance programs, and plan for the future.
This tension is evident with the recent federal pushback against Utah’s Artificial Intelligence Transparency Act (HB 286), a state-level effort to impose transparency requirements, safety planning obligations, and whistleblower protections on large AI developers. The White House’s direct intervention in state policymaking highlights deeper conflicts over legal authority, innovation policy, public safety, and the future of AI governance in the United States.
Why States Are Moving to Regulate AI
With Congress having yet to enact comprehensive federal AI legislation, states have increasingly stepped into the regulatory vacuum. Legislatures across the country have pursued laws addressing algorithmic discrimination, transparency, consumer protection, child safety, and deepfake content. This isn't new. States have always stepped in when federal action stalls, and in AI, that's exactly what's happening.
States including Colorado, California, Texas, New York, Connecticut, and Utah have all pursued different regulatory strategies. Just recently in May 2026, Connecticut lawmakers passed broad AI legislation that includes provisions addressing consumer protection, employment-related AI use, youth safety, and regulatory experimentation through an AI sandbox framework. New York has advanced frontier AI governance legislation requiring large AI model developers to establish safety and security frameworks and report serious incidents associated with advanced AI systems.
In Utah, the Artificial Intelligence Transparency Act (HB 286), introduced in January 2026, would have required frontier AI developers which were defined by size and computing intensity, to publish safety and child protection plans, conduct risk assessments for frontier models, and establish whistleblower protections for employees raising safety concerns. The bill passed unanimously out of committee, but after direct pressure from the White House, the enacting clause was killed in March 2026.
Colorado’s AI Act similarly requires developers and deployers of high-risk AI systems to take steps to prevent unlawful discrimination. Other states, including New Jersey, have enacted laws targeting deceptive AI-generated media, while California has advanced transparency and reporting requirements for advanced AI systems.
Supporters of state-level AI regulation argue that states are responding to real-world harms that organizations and citizens are already experiencing like algorithmic bias, deceptive AI content, and insufficient transparency around automated decision-making, and that waiting for federal action means waiting for protections that may never arrive.
Why the Federal Government Is Pushing Back
The Trump administration has signaled strong opposition to state-by-state AI regulation, favoring a unified national policy framework instead.
On December 11, 2025, President Trump signed an executive order titled Ensuring a National Policy Framework for Artificial Intelligence (Executive Order 14365). The order frames AI governance as both a national security and economic competitiveness issue, arguing that a patchwork of state laws burdens interstate commerce, undermines innovation, and creates conflicting compliance obligations for organizations operating nationally. The administration’s stated position is that U.S. AI companies “must be free to innovate without cumbersome regulation.”
To act on that position, the order directs the Department of Justice to establish an AI Litigation Task Force to challenge state AI laws deemed inconsistent with federal policy, including on constitutional grounds. It also directs the Commerce Department to evaluate existing state laws and flag those considered “onerous,” and authorizes agencies to consider conditioning certain federal funding on states not enacting or enforcing conflicting AI laws.
The Utah episode illustrated just how directly the administration is willing to intervene. According to reporting by Axios, the White House Office of Intergovernmental Affairs sent a letter on February 12, 2026 to Utah Senate Majority Leader Kirk Cullimore Jr. stating the administration was “categorically opposed” to HB 286 and considered it “an unfixable bill that goes against the Administration’s AI Agenda.” White House officials also held multiple conversations with the bill’s sponsor urging him not to move it forward. The pressure worked. The enacting clause was killed in March 2026.
That outcome is significant. HB 286 was narrowly focused on transparency and child safety, areas the December executive order explicitly suggested would be protected from federal preemption. The administration’s willingness to oppose it anyway suggests that even narrowly scoped state AI laws may be subject to federal scrutiny.
In March 2026, the White House released a legislative blueprint for a national AI policy framework, calling on Congress to adopt a federally unified approach centered on preempting state AI laws and a “light-touch” regulatory regime. Whether Congress acts on that blueprint remains to be seen.
The Federalism Question Behind AI Governance
At its core, this is a question about authority: who gets to set the rules for AI when Congress hasn't.
The current debate reflects genuinely competing governance priorities. Advocates for federal uniformity argue that a single national standard simplifies compliance, supports innovation, and eliminates the operational uncertainty created by 50 divergent state requirements. Supporters of state-level regulation argue that states can respond more quickly to emerging harms, that local communities should retain influence over technology governance, and that state experimentation may ultimately inform better national standards.
A unified national standard has real appeal. Dealing with 50 different state requirements is genuinely complicated, and most healthcare organizations don't have the resources to track a different set of rules for every state they operate in. But the current federal approach is built primarily around protecting the ability of AI companies to innovate quickly and with minimal restriction. That is a legitimate priority, but it is not the same as telling healthcare organizations what responsible AI use actually looks like. For leaders trying to make good decisions about AI in clinical settings, a framework that says "move fast and we'll sort out the details later" is not much of a framework at all, especially for a high risk industry like healthcare.
Why This Matters for Healthcare Organizations
So what does that mean for your organization?
Healthcare systems increasingly rely on AI-assisted tools for clinical documentation, medical imaging support, predictive analytics, patient triage, revenue cycle management, prior authorization workflows, and operational forecasting. Many of these tools are deployed across organizations operating in multiple states.
If state AI laws continue diverging, your organization may face inconsistent requirements across jurisdictions, different transparency disclosure standards, different bias auditing expectations, vendor accountability frameworks, and different patient notification obligations. An AI system considered compliant in one state may require additional safeguards or different documentation in another. Procurement teams may need to evaluate vendors differently depending on where your organization operates. Governance committees may find themselves overseeing compliance across overlapping and sometimes conflicting regulatory frameworks.
The uncertainty also makes long-term planning harder. You may be facing decisions about which AI tools to invest in, how to deploy them, and how to train your staff before anyone has told you what the rules are going to be.
That is the reality right now, and it is not going to resolve quickly.
The Governance Challenge for Healthcare Leaders
What you can rely on in the meantime is that some things will be required regardless of how this plays out. Healthcare has always demanded accountability for clinical decisions, clear documentation, and defined oversight. AI doesn't change that standard. It raises it. Building a governance structure now, starting with how you evaluate tools, who is responsible for overseeing them, and how you monitor their performance over time, isn't about getting ahead of regulation. It's about meeting the standard of responsible practice your organization already holds itself to.
That work doesn't have to wait. If you're not sure where to start, that's what we do at Wayde AI.
What Healthcare Leaders Should Do Now
Here is where to start.
Build governance before you scale. Don't wait until AI is embedded across your organization to ask who is responsible for overseeing it. Start with one use case, define who evaluates it, who approves it, and who monitors it over time. That structure is much easier to build before AI is everywhere than after.
Ask your vendors harder questions. When an AI vendor can't clearly explain how their model works, how bias is managed, or how they're preparing for regulatory changes, that's important information. A vendor that can't answer those questions today will be a problem when requirements tighten.
Keep an eye on your states. If you operate across state lines, what's compliant in one state may not be in another. That gap is already real and likely to grow. Someone in your organization needs to watch it.
Plan for oversight from multiple directions. Federal, state, and organizational requirements are all likely to evolve on different timelines. Build your governance program to be flexible rather than fixed to any one set of rules.
Conclusion: AI Governance Will Likely Remain Multi-Level
The federal-state debate over AI regulation is not going to resolve cleanly or quickly. For healthcare organizations, waiting for that resolution before building governance is not an option. The standards that matter most, accountability for clinical decisions, clear oversight, and defined responsibility, are not new. They predate AI. What AI does is raise the stakes for meeting them.
The leaders who navigate this well won't necessarily be the ones who had the clearest regulatory roadmap. They'll be the ones who built responsible practices early and stayed close to the question of what's best for their patients.
Staying informed doesn’t have to mean hours of reading. The Wayde AI Brief is a short weekly intelligence brief for healthcare and mental health leaders navigating real-world AI adoption, governance, and risk. Subscribe for free.
Frequently Asked Questions
Why are states creating their own AI laws? Because federal legislation hasn't arrived. With more than 1,000 AI-related bills introduced across U.S. states and territories in 2025 alone, legislatures are responding to real concerns about algorithmic bias, consumer protection, transparency, and child safety. States are doing what they have historically done when federal action is slow: stepping in on their own.
Why is there a push for a single national AI framework? The argument for federal uniformity is straightforward: 50 different state requirements create real compliance burdens, particularly for organizations operating across state lines. A single national standard would simplify that. The tension is that the current federal approach prioritizes innovation and speed over accountability, which leaves healthcare organizations without much guidance on what responsible use looks like.
How could fragmented regulation affect healthcare organizations? If you operate across multiple states, what's compliant in one state may not be in another. Requirements around transparency, documentation, vendor oversight, and patient notification may differ depending on where you operate. That makes vendor evaluation harder, governance more complex, and long-term planning harder when you don't yet know which rules will apply where.
Should we wait for federal regulation before building a governance program? No. Build it now. The organizations that struggle most are the ones where AI spread through workflows before anyone established who was responsible for overseeing it. Start with clear criteria for evaluating tools, defined accountability, and a process for monitoring performance over time. That foundation holds regardless of what regulation ultimately looks like.
About the Author
Dr. Ernest Wayde is the Founder and Principal of Wayde AI, a healthcare AI ethics consulting firm. He works with healthcare and behavioral health organizations on responsible AI adoption, governance, risk management, and implementation strategy.
References
[1] White House. Ensuring a National Policy Framework for Artificial Intelligence (Executive Order 14365).. White House Presidential Actions, December 11, 2025. https://www.whitehouse.gov/presidential-actions/2025/12/eliminating-state-law-obstruction-of-national-artificial-intelligence-policy/
[2] Axios. Scoop: White House pressures Utah lawmaker to kill AI transparency bill.. Axios, February 15, 2026. https://www.axios.com/2026/02/15/white-house-utah-ai-transparency-bill
[3] Utah Legislature. H.B. 286 Artificial Intelligence Transparency Amendments.. Utah Legislature 2026 General Session, 2026. https://le.utah.gov/~2026/bills/static/HB0286.html
[4] Mintz. White House AI Policy Update: Utah AI Bill Opposition, Tech Corps Launch, and State of the Union Data Center Announcements.. Mintz AI: The Washington Report, February 27, 2026. https://www.mintz.com/insights-center/viewpoints/54731/2026-02-27-white-house-ai-policy-update-utah-ai-bill-opposition
[5] Sullivan & Cromwell. Trump Administration Releases National Policy Framework on Artificial Intelligence.. Sullivan & Cromwell, March 20, 2026. https://www.sullcrom.com/insights/memo/2026/March/White-House-Releases-National-Policy-Framework-AI
[6] Colorado General Assembly. (2024). SB24-205: Consumer protections for artificial intelligence. https://leg.colorado.gov/bills/sb24-205
[7] Office of Governor Gavin Newsom. (2025, September 29). Governor Newsom signs SB 53, advancing California’s world-leading artificial intelligence industry. https://www.gov.ca.gov/2025/09/29/governor-newsom-signs-sb-53-advancing-californias-world-leading-artificial-intelligence-industry/
[8] Texas Legislature. (2025). Texas Responsible Artificial Intelligence Governance Act (HB 149). https://capitol.texas.gov/tlodocs/89R/analysis/html/HB00149S.htm
[9] Office of Governor Kathy Hochul. (2025, December 19). Governor Hochul signs nation-leading legislation to require AI frameworks for AI frontier models. New York State Governor’s Pressroom. https://www.governor.ny.gov/news/governor-hochul-signs-nation-leading-legislation-require-ai-frameworks-ai-frontier-models
[10] George, B., Gressel, A. R., Milaninia, N., Frant, N., Stanislawski, A., & Razipour, K. (2026, May 5). Connecticut poised to enact one of the nation’s most comprehensive AI laws. Freshfields Bruckhaus Deringer LLP . https://www.freshfields.com/en/our-thinking/blogs/a-fresh-take/connecticut-poised-to-enact-one-of-the-nations-most-comprehensive-ai-laws-102mrpv1
[11] New Jersey Legislature. (2025). P.L. 2025, c.040 (A3540): Establishes criminal penalties for production or dissemination of deceptive audio or visual media. https://pub.njleg.gov/Bills/2024/PL25/40_.HTM
[12] Mayer Brown. President Trump Issues Executive Order on 'Ensuring a National Policy Framework for Artificial Intelligence.'. Mayer Brown, December 23, 2025. https://www.mayerbrown.com/en/insights/publications/2025/12/president-trump-issues-executive-order-on-ensuring-a-national-policy-framework-for-artificial-intelligence
[13] Sachs, B. (2025, December 18). More than 1,000 AI bills later, here’s what states are actually doing with artificial intelligence. American Enterprise Institute . https://www.aei.org/technology-and-innovation/more-than-1000-ai-bills-later-heres-what-states-are-actually-doing-with-artificial-intelligence/
Comments