The Trump administration on Friday released a four-page legislative blueprint laying out its vision for national AI regulation, urging Congress to create a single federal framework that would preempt the growing patchwork of state-level AI laws.
The document, formally titled the National Policy Framework for Artificial Intelligence: Legislative Recommendations, arrives as state legislatures across the country race to regulate artificial intelligence on their own terms — with over 1,500 AI-related bills already introduced in 45 states during the 2026 session alone.
Six Guiding Principles
The framework is organized around six core priorities for lawmakers: protecting children online, keeping electricity costs in check as data centers proliferate, safeguarding intellectual property rights, preventing AI-driven censorship, educating Americans on how to use AI responsibly, and removing barriers to innovation.
"Congress should preempt state AI laws that impose undue burdens to ensure a minimally burdensome national standard consistent with these recommendations, not fifty discordant ones," the framework states. States like Washington have already enacted their own AI regulations — here is what Washington's AI laws mean for businesses operating there today.
Child Safety and Consumer Protection
On child protection, the blueprint calls for parental controls over accounts and devices to safeguard children's privacy. It also supports features designed to combat potential sexual exploitation and self-harm facilitated by AI systems.
The framework recommends that Congress address the use of AI replicas that simulate a person's likeness or voice — a growing concern as deepfake technology becomes more accessible.
Energy and Infrastructure
Recognizing the massive energy demands of AI infrastructure, the document calls on Congress to streamline permitting so data centers can generate their own power on-site. It also proposes codifying a requirement that tech companies pay for their increased energy consumption rather than passing costs on to ratepayers.
Copyright and Innovation
In one of its most closely watched positions, the administration stated it "believes that training of AI models on copyrighted material does not violate copyright laws," while acknowledging that "arguments to the contrary exist" and supporting the courts in resolving the matter.
To encourage experimentation, the framework proposes establishing "regulatory sandboxes" that would allow developers to test AI systems under relaxed rules.
Limits on Preemption
Notably, the administration carved out several exceptions. It does not seek to preempt state enforcement of general laws against AI developers for fraud, consumer protection, or child safety. Local authorities would retain control over data center siting decisions and how states procure AI tools for education and law enforcement.
What Comes Next
The administration called on Congress to convert the framework into legislation "in the coming months" — a timeline that puts pressure on lawmakers to act before the 2026 midterm elections shift political dynamics. Whether the blueprint can bridge the gap between pro-innovation and safety-focused factions in Congress remains an open question.
