On Friday, Donald Trump’s administration presented Congress with its proposals for regulating artificial intelligence, urging lawmakers to curb state-level initiatives and adopt a максимально restrained approach to regulation.
The proposal comes in the form of a brief four-page document that seeks to limit states’ ability to set their own rules for the technology. In effect, it revives a longstanding dispute between the federal government, Congress, and regional authorities over how AI policy should be structured.
The document states explicitly: “Congress should preempt state AI laws that impose undue burdens in order to ensure a single national standard with minimal restrictions aligned with these recommendations, rather than fifty disparate regimes.”
Among the proposals are measures to regulate the use of digital replicas that reproduce a person’s likeness or voice, as well as to enshrine in law Trump’s pledge to require technology companies to cover the resulting increase in energy consumption. The document also calls for the creation of so-called “regulatory sandboxes,” where developers would be able to test AI under relaxed rules.
Particular attention is given to child safety online. The document states: “AI services and platforms should take steps to protect children, while also providing parents with tools to oversee the digital environment and the upbringing of their children.”
These proposals are expected to serve as the basis for Republican initiatives in Congress. However, key disagreements—including over federal preemption of state laws, copyright, and child protection—remain unresolved and have for years stalled decision-making. Efforts to curtail state authority, in turn, are likely to face resistance from local politicians in both parties.