White House Urges Congress to Override State AI Laws as Debate Over Regulation Intensifies

BY COMFORT OGBONNA

The White House said Friday that Congress should move to override state laws regulating artificial intelligence that the administration believes are overly burdensome, presenting a broad framework for federal legislation designed to address concerns about the rapidly growing technology while still encouraging innovation and economic growth.

According to the proposal, federal lawmakers should establish a national approach that prevents states from imposing their own complex and potentially conflicting AI regulations. Officials argue that a patchwork of state laws could slow technological progress, create compliance challenges for companies, and weaken the United States’ ability to compete globally in the development of artificial intelligence.

The framework outlines several key principles intended to guide congressional action. Among the priorities highlighted by the administration are protecting children from harmful uses of AI technologies, preventing a surge in electricity costs tied to the rapid expansion of data centers and computing infrastructure, respecting intellectual property rights, preventing the use of AI tools for censorship, and improving public understanding of how to responsibly use the technology.

Republican leaders in the House quickly signaled their support for the proposal, saying they are prepared to work with lawmakers from both parties to craft legislation based on the administration’s recommendations. However, turning the framework into law is expected to be difficult. Any bill would require support from Democrats in the Senate, where divisions over how strictly artificial intelligence should be regulated remain significant.

The administration’s announcement arrives at a time when several state governments have already moved ahead with their own rules governing AI. At the same time, civil liberties organizations and consumer protection groups have increased pressure on policymakers to impose stronger oversight on the rapidly advancing technology.

Technology companies and the White House have argued that allowing states to independently regulate artificial intelligence could lead to a confusing mix of rules that would hinder innovation and discourage investment in the sector. In December, former President Donald Trump signed an executive order aimed at preventing states from creating their own regulatory frameworks for AI development.

Despite the political challenges, the administration’s proposal appears designed to appeal to lawmakers across the political spectrum by addressing concerns that have gained bipartisan attention. These include worries about the potential dangers of AI chatbots interacting with children, as well as the growing electricity demand required to power massive data centers that support advanced AI systems.

Neil Chilson, a former chief technologist at the Federal Trade Commission who now leads AI policy at the Abundance Institute, said the proposal seems aimed at building broader political consensus.

“It covers basically all the key sticking points I think that might stop an AI bill from moving through Congress,” Chilson said. “It reads to me as an attempt to build a larger tent, even if it doesn’t give everybody everything that they want.”

Still, the framework has already drawn criticism from some Democrats who argue that it does not go far enough in holding technology companies accountable for potential harms caused by artificial intelligence.

Representative Josh Gottheimer of New Jersey said the proposal fails to provide meaningful protections for consumers and communities.

“It fails to address key issues, including strong accountability for AI companies, under the guise of protecting children, communities, and creators,” Gottheimer said in a statement. “Americans need protection — but this means nothing if we allow the AI industry to be the Wild West.”

The prospects for federal AI legislation may also depend on the positions taken by influential Republicans in the Senate. Senator Marsha Blackburn of Tennessee has introduced her own legislation addressing artificial intelligence and previously played a key role in blocking an earlier effort by Trump to discourage states from regulating the technology.

Blackburn described the administration’s proposal as a potential roadmap and said she welcomed the White House’s participation in ongoing discussions about how to craft legislation that can pass both chambers of Congress.

Several states have already enacted their own AI-related laws. California, Colorado, Texas, and Utah are among those that have adopted measures aimed at regulating how artificial intelligence is used in the private sector.

In Texas, lawmakers from both parties supported a new law that took effect earlier this year requiring government agencies and health care providers to disclose when artificial intelligence is used to interact with consumers or answer questions. The law also bans the development of AI systems designed to encourage people to commit suicide, harm themselves or others, or engage in criminal activity.

Legal experts say federal legislation modeled on the administration’s framework could override parts of state laws while allowing some provisions to remain in place.

Saurabh Vishnubhakat, a professor at Yeshiva University’s Cardozo School of Law, said such a law could reshape state-level regulations significantly.

“A federal law following this framework could knock out parts of Texas’s AI code while leaving some parts standing,” he said, adding that the political affiliation of state leaders would likely have little impact on whether their laws are overridden.

Colorado’s AI law may also face uncertainty. The measure, passed in 2024, seeks to prevent artificial intelligence systems from discriminating against individuals when making important decisions related to hiring, housing, or medical care. Although the law has been approved, it is not scheduled to take effect until later this year, and lawmakers are considering revisions before it is implemented.

Colorado State Representative Jennifer Bacon said lawmakers must strike a careful balance between encouraging innovation and protecting citizens.

She noted that many voters support technological advancement but do not want the country to abandon safeguards.

California officials also voiced concerns about the administration’s proposal. Governor Gavin Newsom, a Democrat, has taken a mixed approach to artificial intelligence regulation, signing some bills into law while vetoing others.

A spokesperson for Newsom criticized the White House framework, saying it could weaken state efforts to protect residents and consumers.

“Yet again, Donald Trump is trying to gut laws in California that keep our residents safe and protect consumers — a core state responsibility,” spokesperson Marissa Saldivar said.

While the administration argues that federal legislation should limit states from regulating AI development directly, it also says certain state powers should remain intact. According to the framework, states should still be able to enforce general laws that protect children, prevent fraud, and safeguard consumers.

The administration also stated that Congress should not interfere with decisions made by local governments about where to place data centers or other AI infrastructure. Similarly, states should retain authority over how they adopt artificial intelligence tools for use in law enforcement, education, and other government services.

However, the framework argues that states should not be allowed to penalize AI developers for illegal actions carried out by third parties using their technology. It also states that Americans should not face restrictions when using AI for activities that would otherwise be lawful without the technology.

Concerns about the energy demands of artificial intelligence have also become more prominent. As AI systems require enormous computing power, the expansion of data centers has driven up electricity consumption and raised fears about rising energy costs.

In response, the administration has increased pressure on both AI companies and energy providers to address the issue. Earlier this month, several technology firms signed voluntary pledges to develop their own power generation facilities to support AI infrastructure.

Some advocates focused on AI safety argue that the current framework does not adequately address potential long-term risks posed by the technology, including threats to national security, economic stability, and employment.

Brendan Steinhauser, a former Republican strategist who now leads the Alliance for Secure AI, said the proposal does not sufficiently confront the possibility that artificial intelligence could replace large numbers of human workers.

“We have companies that explicitly are hoping to replace human labor,” Steinhauser said. “Tinkering at the edges with upskilling and job training is just not going to make an impact on that. I just don’t think we as a country are taking this seriously enough.”

Another contentious issue addressed in the framework is the relationship between artificial intelligence and copyright law. The document recommends that Congress avoid directly intervening in ongoing legal battles between artists, creators, and technology companies over how copyrighted material is used to train AI models.

The administration’s position is that training artificial intelligence systems on copyrighted material does not necessarily violate copyright law, although it acknowledges that opposing arguments exist. As a result, officials suggest that courts should continue resolving these disputes through existing legal processes.

This stance was welcomed by AI Progress, a trade group representing companies including Amazon, Anthropic, Google, Meta, Microsoft, Midjourney, and OpenAI.

Technology companies are currently facing dozens of lawsuits from writers, publishers, visual artists, and music industry groups who claim their copyrighted works were used without permission to train AI systems capable of generating text, images, and audio.

Courts have largely sided with technology companies so far, ruling that using copyrighted materials to create new outputs may qualify as “fair use.” However, some judges have raised questions about whether the data used to train AI systems was obtained legally.

In one major case last year, a federal judge approved a $1.5 billion settlement between AI developer Anthropic and a group of authors who claimed nearly half a million books had been illegally pirated to train the company’s chatbot.

As artificial intelligence continues to expand rapidly across industries, the debate over how to regulate it — while maintaining innovation and global competitiveness — is expected to remain one of the most complex policy challenges facing lawmakers in the coming years.