What Trump's AI Executive Order (EO 14365) Means for State Insurance Regulation

On December 11, 2025, President Trump signed Executive Order 14365, titled “Ensuring a National Policy Framework for Artificial Intelligence.”¹ The order calls for a minimally burdensome national AI standard and instructs federal agencies to challenge state AI laws that conflict with federal policy.¹ Insurers need to determine how far those challenges can change what state insurance departments require.

The order set four things in motion: a litigation task force, a Commerce Department list of offending state laws, a funding cutoff aimed at states that keep theirs, and a Federal Trade Commission statement of when a state AI law is preempted outright. This article follows each of them to where it actually stands, and explains where McCarran-Ferguson sits if any of them ever reaches an insurance rule.

What is Executive Order 14365?

Executive Order 14365, signed December 11, 2025,¹ makes federal AI policy explicit: the United States will pursue a minimally burdensome national framework, and state laws that conflict with that framework are targets for federal challenge. The order names three problems with state-by-state AI regulation: it creates a patchwork of 50 regimes, it may require models to embed ideological bias, and it can regulate beyond state borders.¹

The order creates a federal challenge mechanism while Congress considers national legislation. Its text leaves federal insurance law and McCarran-Ferguson untouched and never mentions insurance. Carriers therefore face more legal uncertainty while their daily compliance obligations remain state-by-state for now.

What does the order direct federal agencies to do?

The order directs the Attorney General to create an AI Litigation Task Force to challenge state AI laws on the grounds that they unconstitutionally regulate interstate commerce, are preempted by existing federal regulations, or are otherwise unlawful in the Attorney General’s judgment.¹ It directs the Secretary of Commerce to publish, within 90 days, an evaluation identifying state AI laws that are “onerous” or conflict with federal AI policy.¹ And it puts money behind both: the same 90-day clock required a policy notice making states with those identified laws ineligible for broadband non-deployment funds.¹ Its one worked example is Colorado, described only as “a new Colorado law banning ‘algorithmic discrimination’” that “may even force AI models to produce false results.” The text omits a bill number; observers supplied the SB 24-205 identification.¹

Only the first of those is finished. The Attorney General established the Task Force on January 9, 2026, inside the thirty days the order allowed.² The Commerce evaluation was due March 11, 2026, and has not been published. That absence matters more than it looks, because the order chains the rest of its machinery to it. The funding cutoff bites only against states “identified pursuant to section 4.” The FCC proceeding on a federal AI disclosure standard does not begin until ninety days after the same evaluation appears.¹ More than four months past the deadline, the lever and the clock are both still waiting on a document nobody has seen.

Can Executive Order 14365 preempt state insurance regulation?

Federal preemption of state insurance regulation is not automatic. Since 1945, the McCarran-Ferguson Act has said that no Act of Congress shall be construed to “invalidate, impair, or supersede” a state law regulating insurance.³ The exception is narrow: a federal statute that “specifically relates to the business of insurance.”³ This is reverse preemption. It turns the usual supremacy rule around and makes state insurance law the default. Congress passed it after the Supreme Court held that insurance transactions could be regulated as interstate commerce, ceding primary regulatory authority to the states in the same breath.⁴

Executive Order 14365 is a presidential directive, while McCarran-Ferguson speaks to Acts of Congress. Reading the two together, the insurance group at Eversheds Sutherland concludes that Washington would need insurance-specific legislation to displace state insurance regulation, and that the Act will otherwise block federal attempts to challenge state insurance laws.⁵ The same analysis carries a caveat carriers tend to drop. The shield follows the business of insurance. Marketing, fraud detection, procurement, or human-resources uses may fall outside it even when the user is an insurer.⁵ The resulting fight belongs in litigation and legislation; the order itself does not produce an immediate federal takeover of insurance AI rules.

A general federal AI law, or a general executive order about AI, does not automatically displace a state rule governing the business of insurance. Congress could write an insurance-specific statute, but Executive Order 14365 contains no such language and never mentions McCarran-Ferguson.¹

This is the carrier’s ambiguity to manage. The order creates pressure, headlines, and litigation risk, but it does not, by its own terms, wipe out Colorado’s SB 26-189, New York’s Circular Letter No. 7, or the NAIC Model Bulletin’s adoption in 24 states and the District of Columbia.⁶

What did the NAIC say in response?

The NAIC issued a statement on December 16, 2025, expressing deep concern over the order.⁷ The association, which represents chief insurance regulators in all 50 states, the District of Columbia, and the U.S. territories, urged the administration to reconsider and, at a minimum, to “affirm state regulation of AI in the business of insurance.”⁷

The NAIC’s position is not just institutional self-interest. State insurance regulators have argued that the order could implicate “routine analytical tools insurers use every day” and stop regulators from addressing risks in rate setting, underwriting, and claims processing, in the statement’s own words, “even when no true AI is involved.”⁷ If the federal government can preempt state AI rules, it may also be able to preempt state rate regulation, market conduct oversight, and consumer protection enforcement.

The NAIC’s response matters because it signals that state insurance departments are unlikely to stop enforcing their existing AI rules while waiting for federal courts to resolve the preemption question. Carriers should expect state exams, market conduct inquiries, and AI governance attestations to continue on schedule.

What federal preemption of state insurance AI looks like now

The order opened four routes to preemption. All four have moved since December, and none has arrived at insurance.

Path one: litigation, and it has started. On April 9, 2026, xAI sued Colorado Attorney General Philip Weiser in federal district court in Denver to strike down SB 24-205.⁸ On April 24, the United States joined the case as a plaintiff-intervenor.⁹

The form of that move is worth stating precisely, because it is narrower than the headline. The Justice Department did not sue Colorado. It attached itself to a case xAI had already filed, using a provision of the Civil Rights Act of 1964 that gives the government an unconditional right to intervene in equal-protection cases once the Attorney General certifies the case is of general public importance.⁹ The court granted the motion the day it was filed.⁸ The government did not choose the defendant, the statute, or the timing.

What it pleaded is narrower still. The complaint in intervention runs to two counts, both under the Equal Protection Clause.¹⁰ There is no preemption count and no Commerce Clause count, which are the grounds the order gave the Litigation Task Force. That task force has filed no case of its own.²

Three days after the intervention, the Colorado Attorney General agreed not to enforce SB 24-205 until fourteen days after the court rules on a preliminary injunction.⁸ Colorado then signed SB 26-189 on May 14, 2026; it replaces that statute from January 1, 2027.¹¹ The detailed transition and the new law’s insurance deference belong to our Colorado SB 26-189 analysis.

The case still does not challenge an insurance rule. Both complaints address Colorado’s general-purpose statute, and their counts are constitutional: xAI pleads the First Amendment, the dormant Commerce Clause, due process vagueness, and equal protection; the United States pleads equal protection alone.⁸¹⁰ Neither pleading attacks an insurance department bulletin, a rate or form rule, or an AI provision in an insurance code. Because McCarran-Ferguson is addressed to Acts of Congress, these constitutional claims leave the reverse-preemption question unanswered.³

Path two: legislation, and there is now a text. The order asked the White House to prepare a legislative recommendation for a uniform federal framework preempting conflicting state laws.¹ That document arrived on March 20, 2026, as A National Policy Framework for Artificial Intelligence: Legislative Recommendations.¹² It is a set of proposals to Congress. It binds nobody.

The framework’s silences are the part that reaches insurance. It never mentions insurance, insurers, financial services, or McCarran-Ferguson.¹² It asks Congress not to create a new federal AI regulator and to work through “existing regulatory bodies with subject matter expertise” instead, and it asks that any national standard not preempt states’ traditional police powers or their laws of general applicability.¹² Both lines point toward leaving sectoral regulators where they are. Neither says so.

Path three: money. Section 5 directs the Commerce Department to issue a policy notice making states with identified onerous AI laws ineligible for BEAD broadband non-deployment funds, to the maximum extent federal law allows.¹

It has not bitten, for the reason that governs most of this order: the ineligibility attaches only to states “identified pursuant to section 4,” and no section 4 evaluation exists.¹ The policy notice was itself due March 11, 2026, and has not issued. The order’s other funding provision is softer by its own terms. Agencies must assess whether they may condition discretionary grants on states not enforcing conflicting AI laws, which is an instruction to consider rather than to act.¹

Path four: the FTC, and it is the only route with a document on it. Section 7 put the Commission on the same ninety-day clock and told it to explain when state laws requiring alterations to an AI model’s truthful outputs are preempted by the FTC Act’s ban on deceptive practices.¹ What reached the Federal Register on July 7, 2026 is a proposal rather than a finished statement, with comments open until July 31.¹³ Insurance is absent from the proposal. The document says that state law is impliedly preempted where it conflicts with a federal regulatory scheme, and that a state law compelling an AI firm to deceive consumers conflicts with section 5.¹³

That theory is worth carriers’ attention for a reason the AI framing hides. It rests on an Act of Congress, which is the only kind of instrument McCarran-Ferguson was written to answer. And McCarran-Ferguson names the FTC Act in particular. The 1945 statute lets the Sherman, Clayton, and FTC Acts reach the business of insurance only “to the extent that such business is not regulated by State law.”³ Rate and form regulation is the clearest case of business regulated by state law there is. So the shield does its intended work here in a way it may not against a constitutional claim. No one has had to raise it yet.

Across all four routes, the timeline extends across years. Colorado’s SB 26-189 will take effect on January 1, 2027, regardless of what happens in federal court.¹¹

What follows from all four paths is a period of dual compliance rather than a handover. State rules remain operative unless a court grants applicable relief or lawmakers enact a valid federal rule that changes the state-law baseline. Any lower-court order would also have its own scope and could change during appeal.

What should insurers do now?

Carriers need three separate decisions, not a new governance program.

1. Keep following the rules that apply today. EO 14365 is not a reason to pause a Colorado, New York, California, or NAIC-aligned compliance program. The NAIC Model Bulletin remains the multistate baseline where it has been adopted.
2. Track the four federal routes as legal events. A complaint, injunction, final FTC statement, Commerce list, or enacted statute can change the analysis. A speech or policy recommendation cannot do that by itself.
3. Separate insurance activity from adjacent company activity. Underwriting, pricing, coverage, and claims sit closer to the business of insurance than employment, procurement, and general marketing. McCarran-Ferguson’s protection therefore cannot be assumed to cover every AI system an insurer uses.⁵

Where this leaves state insurance regulation

As checked on August 2, 2026, the federal activity has produced an intervention in a private challenge to a general-purpose state statute and a proposed FTC policy statement that does not mention insurance.²¹³ The Task Force has filed no case of its own, Commerce has not published its section 4 list, and no filing tests an insurance rule.¹²¹¹ That is the line to watch: a general state AI statute and an insurance rate or form rule do not present the same McCarran-Ferguson question.

FAQ

Does Executive Order 14365 preempt state insurance AI laws immediately? No. The order directs federal agencies to evaluate and challenge state AI laws, but it is not self-executing. Actual displacement would require successful court relief, an insurance-specific Act of Congress, or another valid federal rule with preemptive effect, and until one of those lands, state insurance AI requirements remain enforceable.

Has the federal government sued a state over an insurance AI rule? No. The Justice Department entered a challenge to Colorado’s general-purpose AI law, SB 24-205, in April 2026,⁹ but as an intervenor in a suit xAI had filed rather than as the plaintiff, and its complaint pleads only equal-protection claims. No federal filing has tested a state insurance regulator’s AI requirement.

Should insurers stop complying with Colorado or New York AI rules because of EO 14365? No. The order has not suspended either state’s insurance requirements. Colorado’s transition is explained in the SB 26-189 guide linked above, and New York’s current underwriting and pricing procedure remains in Circular Letter No. 7.

Footnotes

1. The White House, “Ensuring a National Policy Framework for Artificial Intelligence,” Executive Order 14365, December 11, 2025. Sections referenced above: 1 and 2 (policy), 3 (AI Litigation Task Force), 4 (Commerce evaluation of state laws), 5 (BEAD funding restrictions), 6 (federal reporting and disclosure standard), and 8 (legislative recommendation). The order contains no reference to insurance or to the McCarran-Ferguson Act: https://www.whitehouse.gov/presidential-actions/2025/12/eliminating-state-law-obstruction-of-national-artificial-intelligence-policy/. Official published text: Compilation of Presidential Documents, https://www.govinfo.gov/content/pkg/DCPD-202501186/html/DCPD-202501186.htm ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶ ↩⁷ ↩⁸ ↩⁹ ↩¹⁰ ↩¹¹ ↩¹² ↩¹³ ↩¹⁴ ↩¹⁵ ↩¹⁶

2. Office of the Attorney General, memorandum establishing the AI Litigation Task Force, January 9, 2026: https://www.justice.gov/ag/media/1422986/dl ↩ ↩² ↩³ ↩⁴

3. 15 U.S.C. § 1012(b), McCarran-Ferguson Act: https://www.law.cornell.edu/uscode/text/15/1012 ↩ ↩² ↩³ ↩⁴

4. United States v. South-Eastern Underwriters Association, 322 U.S. 533 (1944), holding that insurance transactions conducted across state lines are interstate commerce: https://supreme.justia.com/cases/federal/us/322/533/. The declaration of policy Congress enacted nine months later is at 15 U.S.C. § 1011: “Congress hereby declares that the continued regulation and taxation by the several States of the business of insurance is in the public interest”: https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title15-section1011&num=0&edition=prelim ↩

5. Eversheds Sutherland, “Trump’s AI Order and McCarran Ferguson: Is the insurance industry fully insulated? Not quite,” February 10, 2026: https://www.eversheds-sutherland.com/en/united-states/insights/trumps-ai-order-and-mc-carran-ferguson-is-the-insurance-industry-fully-insulated-not-quite ↩ ↩² ↩³

6. National Association of Insurance Commissioners, “Implementation of NAIC Model Bulletin: Use of Artificial Intelligence Systems by Insurers,” status as of August 6, 2026 (the reference list on page 2 names 24 states plus the District of Columbia, each with its bulletin number and adoption date): https://content.naic.org/sites/default/files/legal-adoption-map-ai-model-bulletin.pdf. Our state-by-state view of the same map is at /states/. ↩

7. National Association of Insurance Commissioners, “Statement from NAIC on AI Executive Order,” December 16, 2025: https://content.naic.org/article/statement-national-association-insurance-commissioners-naic-ai-executive-order ↩ ↩² ↩³

8. Docket, X. AI LLC v. Weiser, No. 1:26-cv-01515-DDD-CYC (D. Colo., filed April 9, 2026), entries 1, 12, 16, 17 and 24: https://www.courtlistener.com/docket/73171074/x-ai-llc-v-weiser/ ↩ ↩² ↩³ ↩⁴

9. U.S. Department of Justice, Office of Public Affairs, “Justice Department Intervenes in xAI Lawsuit Challenging Colorado’s Algorithmic Discrimination Law,” April 24, 2026: https://www.justice.gov/opa/pr/justice-department-intervenes-xai-lawsuit-challenging-colorados-algorithmic-discrimination ↩ ↩² ↩³

10. Complaint in Intervention of the United States, X. AI LLC v. Weiser, No. 1:26-cv-01515 (D. Colo., April 24, 2026): https://www.justice.gov/crt/media/1437846/dl ↩ ↩²

11. Colorado General Assembly, SB 26-189, section 6-1-1708 (an insurer subject to C.R.S. 10-3-1104.9 is deemed in compliance in the practice of insurance): https://leg.colorado.gov/bills/sb26-189 ↩ ↩² ↩³

12. The White House, “A National Policy Framework for Artificial Intelligence: Legislative Recommendations,” March 2026: https://www.whitehouse.gov/wp-content/uploads/2026/03/03.20.26-National-Policy-Framework-for-Artificial-Intelligence-Legislative-Recommendations.pdf ↩ ↩² ↩³

13. Federal Trade Commission, “Policy Statement Concerning the Suppression of Accuracy in Artificial Intelligence Systems,” 91 Fed. Reg. (July 7, 2026), FR Doc. 2026-13628, published as “Proposed policy statement; request for comments,” comments due July 31, 2026. The statement cites Executive Order 14365 and reasons from implied conflict preemption; the words insurance and insurer do not appear in it: https://www.federalregister.gov/documents/2026/07/07/2026-13628/policy-statement-concerning-the-suppression-of-accuracy-in-artificial-intelligence-systems ↩ ↩² ↩³