I. An Executive Order at Odds with Doctrinal Evolution
In December, President Trump signed an executive order directing the Department of Justice to challenge state artificial intelligence (AI) laws nationwide and encouraging federal agencies to explore withholding federal funds from states with “onerous” AI regulations. Framed as a response to national security threats and global competition, the order warned that state-level regulation risks undermining American “AI dominance” and technological leadership. In January 2026, the Department of Justice announced the creation of an AI Litigation Task Force to operationalize this directive, signaling a coordinated effort to challenge state AI laws through constitutional and preemption-based litigation rather than through congressional action.
What makes the order notable is not merely its ambition, but also the legal assumptions on which it rests. Rather than urging Congress to enact a comprehensive federal framework governing AI, or identifying existing federal statutes that clearly preempt state law, the order relies on executive-led litigation, policy statements, and funding leverage to deter or invalidate state regulation. Central to that strategy is the premise that state AI laws are constitutionally suspect: either because they are preempted by federal authority or because they impermissibly burden interstate commerce. In effect, it seeks to achieve national uniformity not through legislation, but through coordinated executive action designed to pressure courts and agencies into disabling state regulatory efforts.
That strategy presupposes that state AI laws are constitutionally suspect simply by virtue of their diversity. It assumes state regulations fragment national markets and impose impermissible burdens on interstate commerce echoing dormant commerce clause doctrinal arguments seen during an earlier phase of the internet when online services were widely understood as geographically indivisible and compliance with state-specific legal requirements was often technologically infeasible.
Over the past decade, however, courts have revisited those premises. Recent Supreme Court decisions reflect a growing recognition that the technological landscape has changed and that modern platforms can tailor products and services to jurisdiction-specific legal requirements. That recognition has begun to reshape how courts evaluate claims of extraterritoriality and undue burden under the dormant commerce clause.
This essay argues that the executive order’s strategy fails to account for that shift. By relying on increasingly outdated assumptions about technological uniformity, the order risks transforming the dormant commerce clause from a doctrine applied through careful, context-specific adjudication into a blunt deregulatory instrument wielded by the executive branch.
II. The Traditional Pathways for Displacing State Law
Under longstanding constitutional doctrine, state laws are not displaced merely because the federal government prefers a different policy outcome. Preemption ordinarily occurs through one of two mechanisms: Congress enacts legislation that expressly or impliedly preempts state law, or courts determine, through case-specific adjudication, that state law conflicts with valid federal law enacted pursuant to congressional authority. In both cases, congressional action, or at least congressional authorization, is the critical predicate.
This structure reflects the Constitution’s allocation of powers. Congress is vested with authority to regulate interstate commerce and to determine when state regulation must yield to national uniformity. The executive branch, by contrast, is charged with enforcing the laws Congress has enacted. While executive agencies may interpret statutes and exercise enforcement discretion, they do not themselves create preemptive force absent a statutory foundation.
Historically, when new technologies raise novel regulatory questions, states often act first. They do so pursuant to their general police powers, regulating products and services that affect residents’ health, safety, and economic well-being. Congress may later choose to standardize those rules nationally, sometimes preempting state law in the process. Consumer protection, environmental regulation, and financial services oversight all followed this pattern.
Courts have been careful to police efforts by the executive branch to short-circuit this process. When agencies have attempted to displace state law, courts have frequently required a clear statement from Congress authorizing such intrusion into traditional state domains. That restraint reflects both federalism concerns and separation-of-powers principles, and it preserves the judiciary’s role in applying constitutional doctrines to concrete factual settings rather than abstract policy disputes.
In the AI context, Congress has not enacted a comprehensive regulatory framework. In the resulting vacuum, states have adopted targeted measures addressing deceptive practices, discriminatory outcomes, child safety, and transparency in automated decision-making. Whether any particular statute is sound policy or constitutionally permissible depends on context-specific analysis under established doctrines — questions properly resolved through adjudication or legislation, not presumed away through executive policy declaration.
The executive order’s most concrete mechanism is its directive to the Department of Justice to form an “AI Litigation Task Force” charged with challenging state AI laws deemed inconsistent with federal policy, including on the ground that they burden interstate commerce. At first glance, this may appear consistent with the Department’s traditional litigation role.
But the order’s theory of enforcement departs from familiar patterns. Rather than identifying specific federal statutes that state laws allegedly violate, it gestures broadly toward constitutional principles, most notably the Commerce Clause, and towards industry and economic interests. The task force is thus asked to identify and pursue constitutional challenges in the absence of a clear congressional command to displace state law.
Litigation authority, however, does not substitute for substantive lawmaking authority. When the Department of Justice initiates enforcement actions, it must ground those actions in statutes enacted by Congress and operate within the scope of authority those statutes confer. As for regulatory agencies, courts have emphasized that agencies may not manufacture preemption through litigation where Congress has declined to act, particularly in areas of traditional state regulation.
The order also gestures toward using federal funding as leverage, instructing agencies to explore whether discretionary grants could be withheld from states that maintain disfavored AI laws. Here again, executive discretion is bounded. Conditions on federal funds must be authorized by Congress, related to the purpose of the funding program, and articulated with sufficient clarity to permit states to make an informed choice. Executive orders cannot unilaterally impose new funding conditions where Congress has not provided for them.
These institutional concerns arise independently of whether constitutional challenges would ultimately succeed on the merits. They become more pronounced, however, when the order’s litigation strategy is evaluated against the contemporary understanding of the constitutional doctrines it invokes — most notably the dormant commerce clause.
III. From Internet Exceptionalism to Technological Calibration: The Dormant Commerce Clause After Pataki and Paxton
The executive order’s reliance on dormant commerce clause litigation reflects a theory of constitutional conflict that emerged during an earlier era of internet governance. This era was shaped by assumptions about technological indivisibility, geographic indeterminacy, and regulatory impossibility. Those assumptions once animated judicial skepticism toward state regulation of online activity. But they no longer reflect either technological reality or contemporary doctrine.
A. The Era of Internet Exceptionalism: Reno, Ashcroft, and Pataki
In the late 1990s and early 2000s, courts approached the internet as a uniquely borderless medium — one for which traditional jurisdictional distinctions were thought to break down. In Reno v. ACLU, the Supreme Court described the internet as a “vast democratic forum[]” resistant to geographic control, a framing that shaped subsequent constitutional analysis. That skepticism carried forward in Ashcroft v. Free Speech Coalition (Ashcroft I) and Ashcroft v. ACLU (Ashcroft II), where the Court repeatedly emphasized the difficulty — indeed, the near impossibility — of tailoring online content regulation to discrete jurisdictions without burdening lawful speech nationwide.
Lower courts extended this logic beyond the First Amendment. In American Libraries Association v. Pataki, a federal district court invalidated a New York statute regulating the online dissemination of adult content on dormant commerce clause grounds. Crucially, Pataki did not hold that state regulation of online activity was categorically unconstitutional. Rather, it rested on a specific factual premise: that internet publishers lacked any meaningful ability to distinguish in-state from out-of-state users. Because compliance with New York law would require nationwide alteration of content, the court concluded that the statute projected New York’s regulatory preferences beyond its borders and imposed an impermissible burden on interstate commerce.
That reasoning fit comfortably within dormant commerce clause doctrine as it was then understood. Where compliance with a single state’s law effectively dictated conduct nationwide, courts were more inclined to view state regulation as extraterritorial or “clearly excessive” under the balancing framework articulated in Pike v. Bruce Church, Inc. The constitutional weakness was not based on regulatory diversity in the above Commerce Clause doctrinal revolution as applied to the early internet. Rather it was grounded in the inability to comply with one state’s law without restructuring a national market due to technological indivisibility.
B. The Doctrinal Pivot: Pike, Extraterritoriality, and Judicial Restraint
Even during this period, however, dormant commerce clause doctrine did not support a blanket prohibition on state regulation affecting national markets. Under Pike, nondiscriminatory state laws are invalid only where the burden on interstate commerce is “clearly excessive in relation to the putative local benefits.” That inquiry is inherently contextual and fact dependent.
More recently, the Supreme Court has emphasized the limits of judicial intervention under this doctrine. In National Pork Producers Council v. Ross, the Court rejected the argument that California’s animal welfare law was unconstitutional simply because it had substantial upstream effects on a nationally integrated market. The Court made clear that extraterritorial impact alone does not render a state law invalid, and that courts should be wary of converting the dormant commerce clause into a general-purpose tool for deregulation. Regulatory diversity, even when substantial, is a feature of federalism. It is not a constitutional defect.
Taken together, Pike and National Pork Producers underscore a point often obscured in contemporary debates: Dormant commerce clause doctrine does not presume national uniformity. It tolerates, even expects, state variation unless compliance is genuinely infeasible or protectionist in favor of the legislative state’s economy.
C. Technological Change and Judicial Reassessment: Free Speech Coalition v. Paxton
What distinguishes the present moment from the era of Pataki is not merely doctrinal refinement, but judicial recognition that the technological predicates underlying earlier cases have eroded. That recognition is explicit in the Supreme Court’s 2025 decision in Free Speech Coalition, Inc. v. Paxton.
Although Paxton arose under the First Amendment, its reasoning reflects a broader recalibration with clear implications for dormant commerce clause analysis. The Court upheld a Texas age-verification law for certain online pornography sites, rejecting arguments that such requirements necessarily imposed excessive burdens on adult speech nationwide. Central to the Court’s analysis was its acknowledgment that the internet of today is not the internet of Reno or Ashcroft I and II. The Court in Paxton was correct to flag the distinct evolution of the internet relative to the “then-nascent” internet of Reno and Ashcroft. Today, modern platforms possess sophisticated tools including geolocation, identity verification, feature toggling, and jurisdiction-specific access controls that allow them to tailor compliance on a state-by-state basis.
In other words, the technological indivisibility that animated Pataki no longer holds true in 2025 and will not hold true moving forward. Where platforms can segment users, modulate features, and localize compliance without redesigning products nationwide, the constitutional calculus changes. Burdens that once appeared “clearly excessive” under Pike may now be incidental. Regulatory effects that once seemed extraterritorial may now be properly understood as a territorially bounded exercise of a state’s police power.
D. Implications for AI Regulation and the Executive Order
Artificial intelligence services operate squarely within this modern technological environment. AI companies already customize products across jurisdictions in response to global privacy laws, consumer protection regimes, and sector-specific regulations. The existence of divergent state AI laws does not, by itself, compel uniform national redesign. Nor does it inevitably burden interstate commerce in a constitutionally significant way, especially as tech companies see historic valuations.
The executive order’s dormant commerce clause strategy implicitly relies on a Pataki-era understanding of the internet as technologically indivisible, where one state’s laws may have an outsized effect on a smaller network of computers nationwide. But that understanding has been overtaken by both technological reality and judicial doctrine. Courts are increasingly unwilling to presume that regulatory diversity equals constitutional burden, particularly where compliance tools enable localization.
To the extent the executive order treats state AI regulation as presumptively unconstitutional because it risks creating “a patchwork,” it misapprehends the direction of contemporary jurisprudence. The Supreme Court has moved away from internet exceptionalism toward technological calibration, evaluating constitutional claims considering actual capabilities rather than abstract fears of fragmentation.
That shift has significant consequences. Dormant commerce clause challenges to state AI laws will rise or fall based on concrete evidence of infeasibility or excessive burden, not on generalized assertions about innovation, competitiveness, or national dominance. And executive-led efforts to invalidate state law on categorical constitutional grounds are increasingly out of step with how courts now understand both technology and federalism.
The dormant commerce clause prohibits states from enacting laws that discriminate against or unduly burden interstate commerce absent congressional authorization. As courts have repeatedly emphasized, the analysis is highly fact specific. It turns on whether a challenged law is protectionist in purpose or effect, or whether its incidental burdens on interstate commerce are clearly excessive in relation to its local benefits.
There is little evidence that existing state AI laws meet those thresholds. These statutes generally apply evenhandedly to firms providing products or services to in-state residents, regardless of where those firms are headquartered. They do not purport to regulate conduct occurring wholly outside the state, nor do they mandate uniform nationwide design choices as a condition of market access. Instead, they operate in areas like consumer protection, civil rights, and public safety — where state authority has long been recognized.
Against this backdrop, the executive order’s litigation strategy represents a significant departure from dormant commerce clause practice. By directing the Department of Justice to proactively challenge state laws deemed “onerous,” the order appears to assume that regulatory variation itself constitutes a constitutional injury. Dormant commerce clause doctrine, however, does not operate by categorical presumption. It requires judicial evaluation of specific statutes, specific burdens, and specific local justifications.
Using the executive branch to mount a coordinated campaign against state laws on this basis risks converting a judicial doctrine into a generalized deregulatory instrument. That approach bypasses the incremental, case-by-case adjudication courts have long employed and substitutes executive judgment for judicial balancing. In doing so, it threatens to transform a constitutional backstop into a policy lever wielded not by courts or Congress, but by the executive branch acting alone. This becomes a red flag for federalism.
IV. Federalism, Institutional Competence, and the Limits of Executive Preemption
The most consequential implication of the executive order lies not in any single doctrinal claim it advances, but in the institutional model it adopts. Rather than seeking resolution through legislation or adjudication initiated by affected parties, the order positions the executive branch itself as the primary vehicle for suppressing state law across an entire regulatory domain.
Federalism doctrine has long reflected skepticism toward such executive intrusions into areas of traditional state concern absent clear congressional authorization. That skepticism is grounded not only in respect for state sovereignty, but in institutional competence. Congress is constitutionally tasked with determining when national uniformity is necessary and with calibrating the tradeoffs between innovation, economic integration, and local governance. Courts, in turn, resolve concrete disputes based on specific factual records. The executive branch, by contrast, lacks both the democratic legitimacy and the adjudicative posture required to make generalized determinations about the constitutional validity of state law.
The irony is that the order arrives at precisely the moment when states are acting in the manner federalism doctrine has historically encouraged. In the absence of comprehensive federal legislation, and amid Congress’s difficulty reaching consensus on artificial intelligence governance, states have begun to experiment with regulatory approaches to address emerging harms using familiar police powers. That pattern has long characterized the development of consumer protection, civil rights, and public utility regulation.
The constitutional system anticipates this dynamic. States act first; courts review discrete challenges; and Congress responds when national uniformity becomes necessary or desirable. What the system does not contemplate is executive-led nullification that short-circuits this process altogether — particularly where Congress has neither enacted preemptive legislation nor delegated such authority to the executive branch. By displacing state law through litigation directives and funding pressure rather than through statute, the executive order risks altering not only the balance of federal–state authority, but the institutional pathways through which that balance is meant to be maintained.
Conclusion
The executive order directing federal agencies to challenge state artificial intelligence laws reflects a theory of governance rooted in assumptions about both technology and constitutional doctrine that courts have increasingly moved beyond. Modern digital platforms possess the capacity to tailor compliance on a jurisdiction-specific basis, undermining claims that state regulation of online services is inherently extraterritorial or imposes unavoidable burdens on interstate commerce.
Against this doctrinal backdrop, the order’s reliance on executive-led litigation and funding pressure to displace state law absent congressional authorization sits uneasily with established federalism principles. Dormant commerce clause doctrine does not confer upon the executive branch a general power to nullify nondiscriminatory state regulation, nor does it treat regulatory diversity as a constitutional injury. Where national uniformity is warranted, the Constitution assigns that determination to Congress, subject to judicial review but not to unilateral executive action.
As artificial intelligence continues to reshape economic and social life, the central challenge is not whether national coordination is desirable, but how it should be achieved within the constitutional structure. The emerging jurisprudence suggests a clear answer: through legislative judgment, case-by-case adjudication, and institutional restraint rather than through executive assertions of authority that courts themselves have grown increasingly reluctant to endorse.