On March 20, 2026, the White House released its most detailed AI policy statement to date: a national legislative framework calling on Congress to enact comprehensive AI regulation under a single federal standard. For anyone tracking AI governance, whether as in-house counsel, a compliance officer, or a business leader, this document deserves close attention, not for what it promises, but for the regulatory dynamics it sets in motion.
What the Framework Proposes
The framework is organized around seven pillars: protecting children online, safeguarding communities and energy infrastructure, respecting intellectual property rights, preventing censorship, enabling innovation, developing an AI-ready workforce, and critically establishing federal preemption of state AI laws. It focuses heavily on expansion of AI infrastructure and suggests very few mechanisms for limiting AI in any way.
The throughline is a light-touch regulatory posture. The administration opposes the creation of any new federal rulemaking body for AI, preferring instead to channel oversight through existing agencies and industry-led standards. It calls for regulatory sandboxes, streamlined permitting for AI data centers, and resources to help small businesses adopt AI tools. On the enforcement side, it emphasizes combating AI-enabled fraud and protecting minors and other vulnerable populations from exploitation.
The centerpiece, however, is preemption. The framework explicitly urges Congress to override state AI laws that “impose undue burdens,” while carving out limited exceptions for child safety, zoning, and a state’s own procurement decisions.
The Preemption Problem
Federal preemption is a familiar aspiration and a familiar sticking point. The comparison to data privacy is instructive. Despite broad agreement that a national privacy standard would be preferable to the current patchwork, Congress has failed for years to pass comprehensive federal privacy legislation. The result: a growing web of state laws, from California’s CCPA and CPRA to the dozens of state consumer privacy statutes enacted since 2023, each with its own requirements, definitions, and enforcement mechanisms.
AI regulation is tracking in the same direction. States are not waiting on Congress. Legislatures across the country are advancing bills covering automated decision-making, algorithmic discrimination, deepfakes, and AI in employment. The longer federal legislation takes, and given overlapping committee jurisdictions, thin bipartisan consensus, and competing legislative priorities, it could take a while, the more entrenched the state-level landscape becomes.
Implications Worth Watching
Several elements of the framework carry particular significance. On intellectual property, the administration deliberately sidesteps the question of whether training AI models on copyrighted material constitutes fair use, leaving that to the courts. That ambiguity will persist for both AI developers and content creators. Meanwhile, the suggestion that rights holders could collectively negotiate licensing terms with AI providers, shielded from antitrust liability, introduces a novel mechanism worth monitoring.
The administration pays special attention to the inherent tension between intellectual property rights and free expression in the context of content creation in the AI space. Of particular note, the administration teases the possibility of a private right of action for Americans to seek redress for government interference with free expression on AI platforms.
On the workforce front, a concurrent draft bill from Senator Blackburn would require public companies to disclose AI’s effects on their workforce quarterly, potentially going beyond what the White House framework envisions.
The Practical Takeaway
Regardless of whether this framework becomes law, AI governance obligations are coming from Washington, from state capitals, or both. Organizations would be well served to begin building compliance infrastructure now, rather than betting on legislative clarity that may be slow to arrive.
