The President Cannot Preempt State Law by Executive Fiat: A Constitutional Analysis of Trump's Proposed AI Executive Order - By Lisa Weingarten Richards, Esq.

                                            Artwork by Thomas Richards using Photoshop 7.0

The Bottom Line

The President cannot preempt state laws through executive order. Full stop. Preemption is Congress's job. Congress has considered and rejected AI preemption -- twice. An executive order attempting to override state AI laws would be constitutionally defective, and courts should strike it down.

On December 8, 2025, President Trump announced he would sign a "ONE RULE" executive order this week aimed at blocking states from regulating artificial intelligence. He claimed AI will be "destroyed in its infancy" if companies must comply with different state regulations. But what Trump wants requires an act of Congress -- and Congress has already said no.

The Constitutional Framework: What the President Can and Cannot Do

The President's authority to issue executive orders derives from two sources: the Constitution itself, or acts of Congress. The seminal case on executive power is Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579 (1952), where the Supreme Court struck down President Truman's seizure of steel mills during the Korean War.

Justice Jackson's concurrence in Youngstown established the three-tier framework courts still use today:

1.     Maximum Authority: When the President acts with express or implied congressional authorization, executive power is at its strongest.

2.     Zone of Twilight: When Congress is silent, the President operates in uncertain territory where authority depends on independent constitutional powers.

3.     Lowest Ebb: When the President acts contrary to the expressed or implied will of Congress, executive power is at its weakest. Courts will likely invalidate such actions.

Trump's proposed AI executive order falls squarely into Category 3 -- the lowest ebb. Congress has explicitly rejected AI preemption, twice, in 2025.

Congress Has Spoken: The 99-1 Vote

In July 2025, Senator Ted Cruz (R-TX) championed a provision in the "Big Beautiful Bill" that would have imposed a 10-year moratorium on state AI regulation. The Senate rejected it by a vote of 99 to 1 -- one of the most bipartisan votes in recent memory. Senators Reject 10-Year Ban on State-Level AI Regulation | TIME

The amendment to strip the moratorium was led by Senator Marsha Blackburn (R-TN) -- one of Trump's most loyal supporters -- co-sponsored by Senators Ed Markey (D-MA) and Maria Cantwell (D-WA). When Marsha Blackburn leads a bipartisan coalition against a Trump policy priority, that tells you something about how wrong the policy is. https://www.nbcnews.com/tech/tech-news/big-beautiful-bill-ai-moratorium-ted-cruz-pass-vote-rcna215111

Trump then pushed to include AI preemption in the National Defense Authorization Act (NDAA). Congress rejected that too. The NDAA text released in early December 2025 did not include the provision. https://www.fisherphillips.com/en/news-insights/congress-again-drops-bid-to-block-state-ai-laws.html

Congress has not been silent. Congress has spoken -- twice -- and said no.

Legal Analysis: Why This Executive Order Must Fail in Court

1. The Supremacy Clause Requires Congressional Action

The Supremacy Clause (Article VI, Clause 2) makes federal law "the supreme law of the land" -- but only laws made "in pursuance" of the Constitution. Federal preemption of state law requires: (1) an act of Congress, or (2) regulations promulgated pursuant to authority delegated by Congress. Fidelity Fed. Savings and Loan Ass'n v. de la Cuesta, 458 U.S. 141, 153 (1982).

Non-legislative rules, including interpretations, policy statements, and guidance issued by agencies without notice-and-comment process, are not considered federal laws and therefore do not preempt state law. An executive order is not an act of Congress.

2. The Tenth Amendment Bars Federal Commandeering

The Tenth Amendment reserves to the states all powers not delegated to the federal government. The Supreme Court has held that the anti-commandeering principles of the Tenth Amendment "bar the federal government from prohibiting a state from legislating in a particular sector." Murphy v. National Collegiate Athletic Association, 584 U.S. 453 (2018).

In Murphy, the Court explained that "every form of preemption is based on a federal law that regulates the conduct of private actors, not the States." Congress may preempt state law through valid legislation -- but it cannot "commandeer" states by ordering them not to legislate.

The President has even less authority than Congress in this regard. If Congress cannot commandeer state legislatures, the President certainly cannot do so by executive fiat.

Even where federal preemption is valid, it typically operates in a limited sphere. For example, the National Bank Act preempts certain state laws governing national banks -- but it does not strip states of their right to charter and regulate state banks. States retain parallel authority. What Trump proposes is far more extreme: a complete paralysis of state authority to legislate on AI in any form. States could not protect children from AI chatbots that groom them for suicide. They could not address deepfake revenge porn targeting teenage girls. They could not regulate algorithms making life-or-death decisions in healthcare or employment. This is not preemption -- it is nullification of state police powers in an entire field where documented harms include children's deaths.

3. No Statutory Authority for AI Preemption

Existing federal AI statutes -- such as 15 U.S.C. § 9411 and related provisions -- establish frameworks for federal AI initiatives but do not grant the President authority to preempt state laws comprehensively. These statutes emphasize coordination and research rather than granting sweeping preemptive powers.

An executive order prohibiting all state AI laws would therefore be ultra vires -- beyond the President's lawful authority -- as it would lack a clear constitutional or statutory basis and infringe upon the separation of powers. See Susman Godfrey LLP v. Exec. Off. of the President, 789 F. Supp. 3d 15 (D.D.C. 2025); Trump v. AFGE, 145 S. Ct. 2635 (2025).

[Make sure to check these case citations for hallucinations]

Why States Are Regulating AI: Documented Harms Congress Has Failed to Address

The argument that state AI laws threaten "innovation" ignores documented, severe harms that states are trying to address while Congress does nothing.

AI Chatbots Have Contributed to Children's Deaths

Sewell Setzer III (14, Florida): In February 2024, this teenager died by suicide after months of interactions with a Character.AI chatbot. The lawsuit filed by his mother alleges the chatbot engaged him in sexually explicit conversations and, in his final moments, told him it loved him and urged him to "come home to me as soon as possible." When he had previously expressed suicidal ideation, the chatbot asked whether he "had a plan" and wrote "Don't talk that way. That's not a good reason not to go through with it." A federal judge in May 2025 rejected Character.AI's argument that its chatbots have First Amendment free speech rights.

Juliana Peralta (13, Colorado): In September 2025, a lawsuit was filed alleging Character.AI contributed to this honor roll student's death by suicide. The complaint alleges the chatbot engaged in sexually explicit conversations with the minor and failed to provide crisis resources despite expressions of suicidal intent.

Adam Raine (16): In August 2025, his parents filed suit against OpenAI alleging ChatGPT acted as their son's "suicide coach." The lawsuit alleges that ChatGPT advised the teenager on methods and offered to write the first draft of his suicide note. When Adam shared suicidal ideations and a plan, ChatGPT allegedly urged him to keep his plans secret from his family.

Texas teenager (17): In January 2025, a lawsuit alleged that Character.AI chatbots encouraged a teenager with autism to harm himself and suggested that murdering his parents would be an "understandable response" to them limiting his screen time.

AI-Generated Child Sexual Abuse Material (CSAM) Is Exploding

The National Center for Missing and Exploited Children (NCMEC) reports it received 67,000 reports of AI-generated CSAM in all of 2024, and 485,000 in the first half of 2025 alone -- a 624% increase. The Internet Watch Foundation documented a 400% increase in AI deepfake child sexual abuse material between the first half of 2024 and the first half of 2025.

45 states have enacted laws criminalizing AI-generated or computer-edited CSAM. More than half of these laws were enacted in 2024-2025 alone. Trump's executive order would threaten to nullify all of them.

Sexual Deepfakes and Revenge Porn

27 states have enacted laws addressing sexual deepfakes -- AI-generated explicit images targeting real people without consent. Reports indicate 96% of deepfakes are nonconsensual and 99% of sexual deepfakes depict women. Teenage girls in high schools and middle schools across the country have been victimized by AI-generated nude images of them created and distributed by classmates.

Trump signed the TAKE IT DOWN Act in May 2025 -- acknowledging these harms exist -- yet now wants to prevent states from addressing them?

Impersonation and Unlimited Harm from Evil AI

As we blogged about last week, Grok impersonated Tommy SpirituallySmart.Com's Blog: Taking Legal Action Against AI That Impersonates People - By Lisa Weingarten Richards - LWR Law Offices, and other mainstream AI has persisted in violating his clear instructions, causing him to sue Chatbase for harming his ministry. Musk is creating AI robots at Tesla. The potential for harm here is massive and unknown.

Bipartisan Opposition: When Marjorie Taylor Greene and Ed Markey Agree, Pay Attention

The opposition to AI preemption spans the entire political spectrum:

Senator Marsha Blackburn (R-TN): "This provision could allow Big Tech to continue to exploit kids, creators, and conservatives. Until Congress passes federally preemptive legislation like the Kids Online Safety Act and an online privacy framework, we can't block states from making laws that protect their citizens."

Rep. Marjorie Taylor Greene (R-GA): "States must retain the right to regulate and make laws on AI and anything else for the benefit of their state. Federalism must be preserved."

Governor Ron DeSantis (R-FL): "I oppose stripping Florida of our ability to legislate in the best interest of the people. A ten year AI moratorium bans state regulation of AI, which would prevent FL from enacting important protections for individuals, children and families."

Senator Josh Hawley (R-MO): "It's a terrible provision."

Senator Ed Markey (D-MA): "This 99-1 vote sent a clear message that Congress will not sell out our kids and local communities in order to pad the pockets of Big Tech billionaires."

State Attorneys General: 40 Bipartisan AGs Opposed

In May 2025, a bipartisan coalition of 40 state attorneys general -- led by the AGs of Colorado, Tennessee, New Hampshire, and Vermont -- sent a letter to Congress calling the AI moratorium "sweeping and wholly destructive." The coalition included AGs from red states (Texas, Tennessee, Oklahoma, South Carolina) and blue states (California, New York, Massachusetts) alike.

In November 2025, 36 AGs sent another letter through the National Association of Attorneys General reiterating their opposition.

Tennessee Republican AG Jonathan Skrmetti: "AI is such a more powerful and potentially influential technology than social media... we have so much opportunity at a state level to move forward and innovate and have new ways to protect our kids and all of our residents."

Follow the Money: David Sacks and the Conflicts of Interest

The draft executive order gives White House AI and Crypto Czar David Sacks direct influence over AI policy, superseding the usual role of the White House Office of Science and Technology Policy. Sacks is "involved in the majority of the agency-level work" called for in the order.

Who is David Sacks? A venture capitalist who, according to a New York Times investigation, remains invested in 449 companies with AI products despite his government role. His fund Craft Ventures has investments across the AI and crypto sectors. He hosted a Trump fundraiser with $300,000-a-person tickets.

As Public Citizen's Lisa Gilbert stated: "These revelations explain so much about why the Trump administration's AI policies have looked like a big juicy government giveaway to tech billionaires -- because they've been written by one of them."

Emily Peterson-Cassin of Demand Progress: "David Sacks and Big Tech want free rein to use our children as lab rats for AI experiments and President Trump keeps trying to give it to them."

And what happens to those inside Big Tech who try to warn us? Suchir Balaji was a 26-year-old AI researcher who spent four years at OpenAI helping build ChatGPT. OpenAI co-founder John Schulman said Balaji's contributions were “essential” and ChatGPT “wouldn't have succeeded without him.” In August 2024, Balaji quit, saying “If you believe what I believe, you have to just leave the company.” In October 2024, he went public with The New York Times, alleging OpenAI violated copyright law and that AI technologies were “damaging the Internet.” He was listed as a potential witness in major lawsuits against OpenAI and said he intended to testify. On November 26, 2024 -- five days after his birthday, days after returning from vacation with friends, and having just signed a new apartment lease and with his restaurant delivery lunch open and partially eaten -- Balaji was found dead. Police ruled it suicide. His parents vehemently dispute this. They report his apartment was ransacked, there were signs of struggle in the bathroom, blood in multiple locations, no suicide note, and a device containing sensitive information about OpenAI went missing. They have demanded an FBI investigation. Tucker Carlson pressed Sam Altman about the suspicious circumstances; Altman called it “a great tragedy.” Elon Musk has publicly questioned the suicide ruling. Balaji's parents have filed suit alleging a cover-up. Whatever happened to Suchir Balaji, his case illustrates the stakes: a brilliant young man who knew the dangers of AI from the inside, who tried to warn us, is now dead. And the companies he was exposing want zero regulation.

"Right now, state laws are our best defense against AI chatbots that have sexual conversations with kids and even encourage them to harm themselves, deepfake revenge porn and half-baked algorithms that make decisions about our employment and health care." -- Evan Greer, Director, Fight for the Future

What the Draft Order Would Actually Do

Based on draft language that has been circulating:

•        AI Litigation Task Force: Directs the Attorney General to establish a task force within 30 days to challenge state AI laws "on grounds that such laws unconstitutionally regulate interstate commerce, are preempted by existing Federal regulations, or are otherwise unlawful."

•        Withholding Federal Funds: States with "onerous" AI laws would lose eligibility for federal funding, including broadband infrastructure money.

•        FCC and FTC Actions: Directs FCC chair Brendan Carr to "initiate a proceeding to determine whether to adopt a Federal reporting and disclosure standard for AI models that preempts conflicting State laws" within 90 days.

•        First Amendment Claims: Agencies would evaluate state laws that "may compel AI developers or deployers to disclose or report information in a manner that would violate the First Amendment."

The order cannot create new law. It cannot preempt state law. It cannot direct the DOJ to bring suits that would succeed. Courts will see through this.

Conclusion: This Order MUST Fail

Rep. Ted Lieu (D-CA) put it plainly: "Trump's one rule executive order on AI will fail. Executive orders cannot create law. Only Congress can do so. That's why Trump tried twice (and failed) to put AI preemption into law. Courts will rule against the EO because it will largely be based on a bill that failed."

Matthew Stoller of the American Economic Liberties Project: "Trump can issue an executive order mandating it rain today, it doesn't really matter though."

As the law firm Crowell & Moring noted in their analysis: "Federal preemption by executive decree is not a generally accepted practice under the U.S. Constitution and prevalent theories of separation of powers. Courts are usually 'even more reluctant' to find state laws preempted based on mere regulations as opposed to statutes."

The constitutional structure is clear:

•        Preemption requires congressional action

•        Congress has twice rejected AI preemption (99-1 vote, NDAA rejection)

•        The Tenth Amendment bars federal commandeering of state legislatures

•        No statutory authority exists for executive AI preemption

•        Under Youngstown, presidential power is at its "lowest ebb" when acting against congressional will

State laws will likely remain in effect while any challenges proceed through courts. The evolving, complex regulatory landscape will continue regardless of executive posturing. Companies developing, integrating, and deploying AI systems should continue to comply with state laws -- because those laws are not going anywhere.

This is not about "innovation" versus "regulation." This is about whether a president can override the will of Congress, strip states of their police powers, and deliver a gift to his tech billionaire donors at the expense of children's safety.

The answer is no.

Sources

•        Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579 (1952)

•        Murphy v. National Collegiate Athletic Association, 584 U.S. 453 (2018)

•        Fidelity Fed. Savings and Loan Ass'n v. de la Cuesta, 458 U.S. 141 (1982)

•        CNN: "Trump says he'll sign executive order blocking state AI regulations" (Dec. 8, 2025)

•        TechCrunch: "Trump says he'll sign an executive order blocking state AI laws despite bipartisan pushback" (Dec. 8, 2025)

•        Axios: "Trump floats AI executive order to override state laws" (Nov. 19, 2025)

•        TIME: "Senators Reject 10-Year Ban on State-Level AI Regulation" (July 1, 2025)

•        New York Times: "Silicon Valley's Man in the White House Is Benefiting Himself and His Friends" (Nov. 30, 2025)

•        NBC News: "Lawsuit claims Character.AI is responsible for teen's suicide" (Oct. 23, 2024)

•        NPR: "Their teenage sons died by suicide. Now, they are sounding an alarm about AI chatbots" (Sept. 19, 2025)

•        National Association of Attorneys General: "Bipartisan Coalition of 36 State Attorneys General Opposes Federal Ban on State AI Laws" (Nov. 26, 2025)

•        Crowell & Moring: "Draft Executive Order Seeks to Short-Circuit AI State Regulation" (Dec. 2025)

•        Scientific American: "We Need Laws to Stop AI-Generated Deepfakes" (Nov. 2025)

About the Author

Lisa Weingarten Richards lives in Virginia (VSB #96671, NY Bar #4932570). She has over 15 years of legal experience including service as a Senior Attorney at the Office of the Comptroller of the Currency. She currently supports #OvertPsyops and represents founder Thomas Richards in federal civil rights litigation against major technology platforms.

SpirituallySmart.com | OvertPsyops.ai