US unveils 'light‑touch' AI plan

The White House published its four‑page "National Policy Framework for Artificial Intelligence" as legislative recommendations on March 20, 2026, under President Donald J. Trump. The document explicitly asks Congress to preempt state laws that regulate how models are developed or that penalize companies for how their models are used, and it instructs lawmakers not to create new federal agencies to police AI. The framework asks Congress to codify the March 4, 2026 "Ratepayer Protection Pledge" and points to commitments by seven companies—Amazon, Google, Microsoft, OpenAI, Meta, Oracle and xAI—to build, buy or pay for generation and grid upgrades so residential ratepayers are not billed for hyperscaler data‑center power. On children’s safety the recommendations call for commercially reasonable, privacy‑protective age‑assurance (such as parental attestation), platform features to reduce risks of sexual exploitation and self‑harm for minors, and an affirmation that existing child‑privacy limits apply to model training data while urging Congress to avoid "open‑ended liability." The administration’s wish list also includes regulatory sandboxes to accelerate AI use cases, streamlined federal permitting for AI infrastructure, grants and tax incentives for small businesses adopting AI, and federal support for skills training plus data collection on AI‑driven job displacement. Legal and policy advisers emphasize the framework is non‑binding and would require congressional legislation to take effect, and several law‑firm analyses warn that the White House’s push for broad federal preemption could face legal limits and challenge from states.