Trump Centralizes AI Regulation, Sidelines States

A Centralized Legislative Framework for AI in the United States

The Trump administration recently unveiled an ambitious legislative framework aimed at establishing a unified policy on artificial intelligence (AI) across the United States. This framework seeks to centralize power in Washington, preempting existing state laws on AI. This move could potentially hinder recent state initiatives that attempt to regulate the use and development of this technology.

In a statement from the White House, it is emphasized that "this framework can only succeed if it is applied uniformly across the United States." According to the administration, a patchwork of conflicting state laws could harm American innovation and compromise the United States' position in the global competition for AI leadership.

The Framework's Goals: Innovation and Centralization

The legislative framework outlines seven main objectives aimed at promoting innovation and the expansion of AI. It proposes a centralized federal approach that would override stricter regulations put in place by certain states. Significant responsibility is placed on parents regarding the safety of children, while expectations for platform accountability remain relatively light and non-binding.

For example, the framework suggests that Congress should require AI companies to incorporate features aimed at "reducing the risks of sexual exploitation and harm to minors." However, it does not specify clear and enforceable measures to achieve this goal.

An Executive Order to Challenge State Laws

This legislative framework follows an executive order signed by Trump three months earlier, directing federal agencies to challenge state laws on AI. This order gave the Department of Commerce a 90-day deadline to compile a list of state laws deemed "burdensome," thereby threatening the eligibility of states for federal funds such as broadband grants. To date, this list has not yet been published.

The order also called for the administration to work with Congress to develop uniform legislation on AI. This vision continues Trump's previous strategy on AI, which prioritized business growth over the establishment of safeguards.

A Minimally Binding National Standard

The new framework proposes a "minimally binding national standard," reflecting the administration's broader push to "remove outdated or unnecessary barriers to innovation" and accelerate the adoption of AI across various sectors. This pro-growth, low-intervention approach is supported by "accelerators" such as David Sacks, the White House AI lead and venture capitalist.

While the framework mentions federalism, exceptions for states are relatively limited, preserving only their authority over general laws such as fraud and child protection, zoning, and the use of AI by states. It draws a strict line against state regulation of AI development itself, which it considers an "intrinsically interstate" issue related to national security and foreign policy.

Protecting AI Developers

The framework also seeks to prevent states from "penalizing AI developers for the illegal conduct of a third party involving their models" — a key liability shield for developers.

This framework lacks gestures toward accountability frameworks, independent oversight, or enforcement mechanisms for potential harms caused by AI. Indeed, the framework would centralize AI policy-making in Washington while reducing the space for states to act as primary regulators of emerging risks.

Reactions and Criticism

Critics argue that states are the laboratories of democracy and have been quicker to adopt laws addressing emerging risks. Notably, New York's RAISE Act and California's SB-53 seek to ensure that large AI companies have and adhere to publicly documented safety protocols.

Brendan Steinhauser, CEO of the Alliance for a Secure AI, stated: "The White House AI lead, David Sacks, continues to serve the interests of large tech companies at the expense of ordinary Americans and workers. This federal AI framework seeks to prevent states from legislating on AI and provides no pathway to accountability for AI developers regarding the harms caused by their products."

Many players in the AI industry celebrate this direction as it offers them more freedom to "innovate" without the threat of regulation.

Teresa Carlson, president of General Catalyst Institute, asserted: "This framework is exactly what startups were asking for: a clear national standard so they can scale quickly. Founders should not have to navigate a patchwork of conflicting state laws that hinder innovation."

Child Safety, Copyright, and Free Speech

The framework has been issued at a time when child safety has become a central point in the AI debate. Some states have aggressively acted to adopt laws aimed at protecting minors and shifting more responsibilities onto tech companies. The administration's proposal takes a different direction, placing more emphasis on parental control than on platform accountability.

"Parents are best positioned to manage the digital environment and education of their children," the framework states. "The administration calls on Congress to provide parents with tools to do this effectively, such as account controls to protect their children's privacy and manage their device usage."

The framework also asserts that the administration "believes" that AI platforms should "implement features to reduce the potential for child sexual exploitation and encouragement of self-harm." While it calls on Congress to require such protections and states that existing laws, including those prohibiting materials of sexual abuse against children, should apply to AI systems, the proposal uses qualifiers such as "commercially reasonable" and does not specify clear prerequisites.

Copyright and Free Speech

Regarding copyright, the framework attempts to find common ground between protecting creators and allowing AI systems to be trained on existing works, citing the need for "fair use." This type of language reflects the arguments that AI companies have made in the face of a growing number of copyright infringement lawsuits concerning their training data.

The main safeguards that the Trump AI framework seems to outline involve ensuring that "AI can pursue truth and accuracy without limitation." It specifically focuses on preventing government censorship rather than on the moderation of platforms themselves.

"Congress should prevent the U.S. government from compelling technology providers, including AI providers, to ban, coerce, or modify content based on partisan or ideological agendas," the framework states. It also instructs Congress to provide a means for Americans to seek legal redress against government agencies that attempt to censor expression on AI platforms or dictate the information provided by an AI platform.

The framework arrives as Anthropic is suing the government for allegedly violating its First Amendment rights after the Department of Defense (DOD) designated it as a supply chain risk. Anthropic argues that the DOD is doing so in retaliation for refusing to allow the military to use its AI products for mass surveillance of Americans or for making targeting and firing decisions in autonomous lethal weapons. Trump has labeled Anthropic and its CEO Dario Amodei as "woke" and "radical leftists."

The language of the framework, which emphasizes the protection of "legal political expression or dissent," appears to build on Trump's previous executive order targeting "woke AI," which pushed federal agencies to adopt systems deemed ideologically neutral.

It is unclear what qualifies as censorship versus standard content moderation, making this type of language potentially challenging for regulators to coordinate with platforms on issues such as misinformation, electoral interference, or public safety risks.

Samir Jain, vice president of policy at the Center for Democracy and Technology, noted: "[The framework] rightly states that the government should not compel AI companies to ban or modify content based on 'partisan or ideological' agendas, yet the administration's 'woke AI' executive order this summer does exactly that."