The Trump administration is close to establishing a national framework for artificial intelligence, according to The Information. The move could reshape AI regulation in the United States and have wider implications for the global technology sector. By shifting AI policy-making to Washington, the proposal would reduce the patchwork of state rules that developers currently have to navigate.

That matters because many of the leading AI technologies are developed by American companies. Rules adopted in Washington would likely affect businesses and regulators beyond U.S. borders as well. One major question remains unresolved: how the U.S. government would treat open-source AI.

Washington wants one rulebook instead of fifty

At the center of the debate is the White House’s National Policy Framework for Artificial Intelligence, published on March 20, 2026. Despite its name, the document is not a regulation or an executive order. Experts at Georgetown University’s Center for Security and Emerging Technologies (CSET) say it should be understood as a legislative proposal for Congress rather than an operative policy.

CSET analysts Mina Narayanan, Jessica Ji, and Vikram Venkatram described the framework as a starting point for negotiations. In their view, it is best read as a way for the administration to state its priorities while leaving Congress to shape the final AI legislation.

The central aim is straightforward: replace multiple state-level AI laws with a single federal standard. The initiative stems from an executive order issued in December 2025, which directed White House officials to develop recommendations for a nationwide AI policy.

The proposal follows limited progress in Congress on measures that would have barred states from enforcing new AI laws or tied some federal funding to state limits on their own AI rules. Those efforts reflected concern that AI developers could otherwise face a growing number of conflicting requirements.

The issue is already real. Colorado’s AI Act has introduced rules aimed at preventing algorithmic discrimination in high-risk AI systems. Texas has reportedly announced its own transparency and consumer protection rules, while California continues to work on legislation involving frontier models and AI-generated content. Although the approaches differ, they contribute to a more fragmented regulatory landscape.

The White House says a unified federal framework would give developers clearer guidance while addressing issues such as child safety, AI-generated deepfakes, and the use of regulatory sandboxes to encourage innovation. If Congress adopts such a framework, companies could spend less time adapting products to different state laws and more time on development.

Trump’s AI strategy goes beyond Congress

Legislation is only one part of the administration’s approach. On June 2, 2026, President Donald Trump signed Executive Order 14409, titled “Promoting Advanced Artificial Intelligence Innovation and Security,” which directs federal agencies to strengthen cybersecurity for advanced AI technologies.

The order calls for an AI cybersecurity clearinghouse, operated with the private sector, to speed up the identification of software vulnerabilities. It also introduces the concept of “covered frontier models,” referring to the most advanced AI systems, and instructs agencies to develop a plan for their safe use.

Together, the legislative framework and the executive order reinforce the administration’s stated goal of preserving U.S. leadership in AI while improving cybersecurity protections.

Open-source remains the open question

The biggest unresolved issue is how open-weight AI models would be handled. The framework addresses topics such as child safety, digital identity, deepfakes, and innovation, but it does not clearly define how open-source AI should be regulated. According to The Information, that is one of the proposal’s main unanswered questions.

The debate comes as governments around the world place increasing emphasis on advanced AI. The International AI Safety Report 2026, led by Turing Award winner Yoshua Bengio and written by more than 100 experts from over 30 countries, says increasingly capable AI systems require stronger governance. The report also warns that open-weight models present unique governance challenges because they can be freely modified and redistributed after release, making consistent safety protections harder to enforce.

For developers, the immediate priority may be regulatory clarity rather than open-source policy itself. A federal regime that overrides state rules could turn a variety of compliance requirements into one national standard, reducing costs for companies operating across all 50 states. At the same time, it would limit states’ ability to build their own AI safeguards before federal rules are finalized.

It remains unclear whether Congress will approve the proposal. Even so, the framework signals that Washington wants to become the primary authority on AI regulation. Because American companies still produce much of the world’s AI technology, the outcome could shape not only U.S. law but also the standards that many developers elsewhere will need to follow.