A four-page document from the White House now sets the terms for every AI regulation fight in Congress this year. Released on March 20, the "National AI Legislative Framework" lays out seven priority areas and a clear philosophy: regulate as little as possible at the federal level, and block states from doing more.
The practical effects, if Congress acts on any of this, will touch every company building or using AI tools.
The Seven Priorities
The framework covers:
- Protecting children and empowering parents - Platforms accessible to minors should reduce exploitation risks, eliminate child user data collection, and give parents better device management tools.
- Safeguarding communities - Data centers should generate their own power. Ratepayers should not subsidize AI infrastructure buildouts. (Trump announced on February 25 that major tech companies agreed to absorb energy surges themselves.)
- Intellectual property and creator rights - This is the most contentious section. The administration states plainly that AI scraping copyrighted material from the internet does not violate U.S. copyright law. At the same time, it asks Congress to develop licensing frameworks so rights holders can negotiate compensation. Courts still get final say on fair use.
- Free speech - AI systems should not enable government censorship or silence lawful political expression.
- Innovation and global dominance - Remove outdated deployment barriers. Create regulatory sandboxes (controlled testing environments where companies can trial AI applications without full regulatory compliance) for developers.
- Workforce development - Expand AI skills training through non-regulatory methods.
- Federal preemption of state laws - The most significant structural proposal. The framework explicitly calls for overriding state AI laws that impose "undue burdens," arguing that "a patchwork of conflicting state laws would undermine American innovation."
No New Regulators, No New Agencies
The framework rejects creating any new federal body to oversee AI. Instead, it maintains the current "sector-specific" approach where existing agencies like the FTC, FDA, and SEC handle AI within their existing jurisdictions. This is a direct rebuke to proposals from some Democrats and EU-style regulation advocates who want a dedicated AI authority.
Regulatory sandboxes are the one proactive idea here. These would let companies test AI applications in a controlled environment with relaxed rules, similar to what financial regulators have used for fintech startups.
Who Wants What
House Republican leaders - Speaker Mike Johnson, Reps. Steve Scalise, Brian Babin, Brett Guthrie, and Jim Jordan - immediately endorsed the framework, pledging to work "across the aisle" on legislation.
The Business Software Alliance welcomed the workforce development and data access provisions. NetChoice director Patrick Hedger praised the light-touch approach, comparing it to the regulatory environment that enabled internet growth in the 1990s.
Not everyone is on board. Americans for Responsible Innovation president Brad Carson criticized the framework for shielding AI developers from liability, warning it creates "open season on the American public."
What This Means for AI Tool Users
Three things to watch. First, the copyright stance is a green light for AI companies training on public web data - expect this to become a major lobbying battleground with publishers and creators pushing back hard. Second, state preemption would kill laws like California's proposed AI safety legislation, creating a single (and lighter) national standard. Third, the "no new agency" position means enforcement will remain fragmented across existing regulators, which historically means slower action.
Passing any of this in a midterm election year is a heavy lift. But this framework now defines the playing field. Every AI bill introduced in Congress will be measured against these seven points, whether lawmakers agree with them or not.