The Federal–State AI Showdown: Why America’s Next Big Tech Fight Runs Through Annapolis

The United States Capitol building superimposed with a digital circuit design, symbolizing the intersection of technology and governance.

By Michael Phillips | November 25, 2025

As 2025 draws to a close, the United States finds itself in the middle of a regulatory standoff that could determine the future of artificial intelligence. With Congress stalled, states sprinting ahead, and the White House weighing federal preemption, the question is no longer whether the U.S. will regulate AI—but who will control the rules: Washington or the states.

More than 1,000 state-level AI bills were introduced in 2025 alone, with 100 enacted across 38 states. Meanwhile, the federal government—despite aggressive moves from the Trump administration—has yet to pass a comprehensive AI law. In the absence of uniform standards, the states have become what Justice Brandeis once called “laboratories of democracy,” testing rules on deepfakes, algorithmic bias, hiring discrimination, and consumer disclosures.

The stakes are enormous. The U.S.–China AI race depends on innovation speed, but so do consumer safety, election integrity, and economic fairness. And now, both sides—federal unifiers and state experimenters—are gearing up for one of the biggest regulatory clashes of the decade.


Federal Push: Centralization, Preemption, and the Battle for Uniformity

The Leaked Executive Order: A Heavy-Handed Attempt to Override States

Between November 19–24, a draft executive order circulated inside the White House titled:

“Eliminating State Law Obstruction of National AI Policy.”

It set off alarms among state leaders and civil liberties groups. Key provisions included:

1. DOJ “AI Litigation Task Force”

A new unit inside the Department of Justice—led by Attorney General Pam Bondi—would sue states whose laws allegedly:

  • burden interstate commerce
  • conflict with federal guidelines
  • “restrict free expression” by requiring AI disclosures

Targets reportedly included California’s SB 53 and Colorado’s SB 24-205—two of the most influential state AI laws.

2. Withholding Broadband Funds

The draft ordered agencies to block BEAD broadband funding from “noncompliant states.” This struck red and blue governors alike as coercive, triggering early talks of 10th Amendment challenges.

3. Federal Agency Preemption

The FCC would develop a national AI disclosure rule to invalidate state-level notification laws.

4. National Labs + “Genesis Mission”

The order folded into the administration’s broader Genesis Mission, which mobilizes national labs to deploy frontier AI models across scientific domains under a “build out, not bailout” framework championed by AI Czar David Sacks.

The Plan Falls Apart

By November 21, internal resistance forced a pause. Senate Republicans—including Josh Hawley and Tom Cotton—warned the White House that conservatives could revolt against federal overreach. By November 24, Reuters reported the administration would pivot toward legislation instead of unilateral action.


The NDAA Becomes the New Battlefield

With the executive order shelved, the administration and House Republicans are turning to the National Defense Authorization Act, the must-pass annual defense bill.

The NDAA’s Proposed AI Preemption Rider

House leaders are exploring a provision that would:

  • bar states from regulating AI development or deployment
  • override existing state AI laws
  • authorize federal standard-setting across the industry

This would amount to blanket federal preemption—something tech companies have been lobbying for intensely.

But Senate Opposition Is Fierce

Sen. Chris Murphy (D-CT) has become the leading opponent, warning:

“This would let an unregulated industry write its own rules.”

Some Republicans are also breaking ranks. Sen. Hawley and Rep. Marjorie Taylor Greene have loudly defended state authority, framing the issue as a fight for federalism.

Odds of Passage? Low but Uncertain

The NDAA’s December timetable leaves little room for negotiation, but the pressure is high. If preemption slides into the final bill, lawsuits from state attorneys general are virtually guaranteed.


States Surge Ahead: 2025’s Patchwork or Progress?

A color-coded map of the United States indicating state-level AI legislation status, with states categorized as 'Proposed Legislation', 'Enacted Legislation', 'Enacted and Proposed Legislation', and 'No Legislation Proposed'.

While Washington debates, states are solving real-world problems right now. And unlike federal proposals, most state laws are targeted—focused on specific risks, not sweeping regimes.

Below is the updated picture for 2025:

California – SB 53 (Transparency in Frontier AI Act)

Effective January 2026. Requires disclosure when interacting with AI and mandates safety reporting.

Focus: Deepfakes, consumer transparency.

Colorado – SB 24-205 (AI Act)

Requires impact assessments for high-risk systems, especially hiring tools.

Focus: Employment discrimination.

New York – RAISE Act

Mandates safety plans and 72-hour incident reporting for frontier AI.

Focus: State agency transparency, risk mitigation.

Utah – HB 452

Requires companies to tell consumers when they are talking to AI.

Focus: Consumer protection.

Texas – Bipartisan Senate Pushback

Texas lawmakers on November 24 urged Sens. Cornyn and Cruz to oppose federal preemption.

Focus: Child safety, consumer rights; defending state sovereignty.

Illinois – Algorithmic Accountability Enforcement

One of the strongest systems for auditing hiring algorithms.

Focus: Bias mitigation.

Oregon – Healthcare AI Restrictions

Restricts AI from using protected medical titles without oversight.

Focus: Patient safety.

Montana – Right to Compute Act

Protects AI infrastructure deployment from local restrictions.

Focus: Data center and energy expansion.

Emerging Trends

  • 19 states advancing algorithmic pricing transparency rules.
  • 4 states enacting digital replica protections (preventing unauthorized AI likeness use).
  • Healthcare AI becoming a major battleground.
  • Election deepfake bans expanding rapidly ahead of 2026 races.

Gov. Ron DeSantis of Florida called the federal preemption push:

“A blatant example of Washington bullying.”

Rep. Marjorie Taylor Greene added:

“States must retain the right to regulate.”

Even blue states like California and New York are preparing for litigation.


Who Wants Preemption? Who Wants State Power?

Pro-Federal: Big Tech + Certain Lawmakers

Tech giants and major venture capitalists argue:

  • The patchwork of state laws raises compliance costs.
  • Uniform standards keep the U.S. competitive with China.
  • Fragmentation will slow innovation and drive investment offshore.

VCs like Andreessen Horowitz have spent heavily to support federal standardization.

Pro-State: Safety Advocates + Bipartisan Lawmakers

Groups like PauseAI and the Torchbearer Community argue:

  • Preemption creates a “regulatory vacuum.”
  • Big Tech wants weaker, not uniform, rules.
  • States are closer to the harms—deepfakes, hiring bias, predatory chatbots.

Sen. Bernie Sanders argues preemption would let:

“Oligarchs enrich themselves on the backs of workers.”

The New Bipartisan Middle

Even MAGA-aligned conservatives are uneasy with federal dominance:

  • Greene, DeSantis, and multiple Texas legislators cite states’ rights.
  • Former Biden antitrust advisor Tim Wu warns that deregulation expands monopolies.

What Happens Next?

The pause of the executive order was not a retreat—it was a reset.

Scenario 1: Preemption Sneaks Into the NDAA

Expect:

  • immediate lawsuits from state AGs
  • federalism fights at the Supreme Court
  • political backlash in both parties
  • Silicon Valley celebrating

Scenario 2: NDAA Fails, States Accelerate

If preemption dies in the Senate, states will:

  • pass more algorithmic transparency laws
  • tighten hiring and healthcare rules
  • expand protection against election deepfakes

This could pressure Congress into action in 2026.

Scenario 3: A Hybrid Compromise

Congress may attempt:

  • a federal floor with room for state experimentation
  • national security carveouts
  • uniform disclosure rules

But no compromise currently has bipartisan momentum.


Conclusion: A Federalism Stress Test for the AI Era

AI is now a defining issue for the economy, national security, and public trust. But the U.S. lacks a coherent national framework. For now, the real innovation—and real governance—is happening in the states.

The next major flashpoint is the NDAA vote in December. If federal preemption language survives the conference committee, it could reshape the balance of regulatory power for decades.

If it fails? Expect 2026 to be the year when states, not Washington, become the true architects of America’s AI future.


Discover more from Maryland Bay News

Subscribe to get the latest posts sent to your email.

Leave a Reply

Discover more from Maryland Bay News

Subscribe now to keep reading and get access to the full archive.

Continue reading