White House AI Proposal Seeks to Override State Laws, Avoid New Regulator

The White Home on Friday launched a sweeping nationwide coverage framework for synthetic intelligence, outlining suggestions to Congress that may set nationwide requirements for AI whereas counting on current federal businesses—moderately than creating a brand new regulator.

The proposal comes as states move ahead with their very own AI legal guidelines, which the Trump administration has criticized as a burdensome “patchwork” of necessities for firms.

“The Trump Administration is dedicated to profitable the AI race to usher in a brand new period of human flourishing, financial competitiveness, and nationwide safety for the American folks,” the White Home stated in a statement. “Attaining these targets requires a commonsense nationwide coverage framework that each permits American business to innovate and thrive and ensures that each one People profit from this technological revolution.”

The framework urges Congress to set nationwide AI guidelines that tackle little one security, innovation, free speech, and mental property, whereas preempting state legal guidelines it views as burdensome. It additionally says these federal requirements mustn’t override states’ current authority to implement legal guidelines on points like fraud, client safety, and little one sexual abuse materials.

The Heart for Democracy and Expertise stated the proposal contains “some sound statements of ideas,” however doesn’t resolve competing priorities.

“Its usefulness to lawmakers is proscribed by its inside contradictions and failure to grapple with key tensions between numerous approaches to vital subjects like children’ on-line security,” CDT Vice President of Coverage Samir Jain stated in a press release shared with Decrypt.

Jain additionally stated the framework contradicted the White Home’s personal place on authorities affect over AI platforms.

“It rightly says that the federal government mustn’t coerce AI firms to ban or alter content material based mostly on ‘partisan or ideological agendas,’ but the administration’s ‘woke AI’ govt order does precisely that,” he stated.

The framework follows earlier efforts by the Trump administration to curb state-level AI regulation. In November, a draft govt order outlined steps to problem state legal guidelines and limit funding to those who enacted legal guidelines that had been seen as contradictory to the order.

Regardless of the administration’s makes an attempt to set a federal normal, states have continued to go their very own measures. In October, California enacted SB 243, which might require AI companion chatbots to determine themselves and limit sure interactions with minors whereas imposing disclosure guidelines on massive builders.

The White Home’s framework additionally stated dad and mom ought to be given extra management over how kids work together with AI methods, and that Congress ought to enact higher protections towards abuse.

“The administration is looking on Congress to offer dad and mom instruments to successfully do this, equivalent to account controls to guard their kids’s privateness and handle their gadget use,” the White Home stated. “The administration additionally believes that AI platforms more likely to be accessed by minors ought to implement options to cut back potential sexual exploitation of kids or encouragement of self-harm.”

The administration additionally stated that whereas it views AI coaching on copyrighted materials as lawful, it believes courts ought to determine the problem, including that Congress “mustn’t take any actions that may influence the judiciary’s decision of whether or not coaching on copyrighted materials constitutes honest use.”

The proposal additionally requires a federal regulation to guard people from unauthorized AI-generated deepfakes, increasing on a bipartisan regulation signed by Trump final 12 months that made non-consensual intimate photos and deepfake porn a federal crime. The brand new framework, nevertheless, comes with exceptions for parody, satire, information reporting, and “different expressive works protected by the First Modification.”

The plan ties AI coverage to infrastructure and financial targets, together with sooner allowing for information facilities and making certain residential electrical energy prices don’t rise on account of AI infrastructure buildout below a proposed “Ratepayer Safety Pledge.” It additionally requires expanded use of on-site and behind-the-meter energy era to help information heart growth and enhance grid reliability, together with incentives to develop AI adoption and entry to federal datasets.

Client advocacy group Public Citizen known as the proposal “a nationwide framework to guard Large Tech on the expense of on a regular basis People.”

“It’s a unprecedented payback to the Large Tech firms which have lined as much as throw pocket change at Trump’s inauguration, and for his ballroom, and for the Melania film, and to settle unhealthy religion lawsuits and extra,” co-president Robert Weissman stated in a statement shared with Decrypt.

Weissman stated the concentrate on preempting state legal guidelines might go away gaps in oversight, arguing that with out new federal requirements, limiting state motion would cut back regulation. He pointed to ongoing state efforts addressing points equivalent to deepfakes, AI companions, and algorithmic decision-making.

“It is a disgraceful proposal that, fortunately, will probably be useless on arrival in Congress,” Weissman stated. “It does, nevertheless, present but once more that Donald Trump aligns his pursuits with the most important firms and the billionaire class, not these of the American folks.”