Trump Administration Changes How the U.S. Regulates Artificial Intelligence

Since taking office in January 2025, the Trump administration has shifted the federal government's approach to regulating artificial intelligence. The new direction focuses on encouraging companies to develop AI quickly and freely, in contrast with the previous approach, which emphasized safety concerns and careful oversight.

The most significant change came on January 20, 2025, when President Trump cancelled the Biden administration's major AI safety order from October 2023. That earlier order had required companies developing advanced AI systems to report on their work, test their systems for risks, and follow specific safety guidelines.

A New Direction: Speed Over Caution

Three days after taking office, Trump signed a new executive order titled "Removing Barriers to American Leadership in Artificial Intelligence." This order signals a clear break from the previous approach. Instead of requiring safety checks, the new framework aims to remove rules that slow down AI development.

The administration has continued this direction with two more orders. One focuses on teaching young Americans about AI. Another directs research into using AI to find treatments for childhood cancer and comes with an additional $50 million in funding. Rather than creating broad safety rules, this approach targets specific problems the government wants AI to solve.

In July 2025, the Trump administration released a plan to simplify environmental rules around AI and to help American allies obtain advanced AI technology. The Commerce and State departments will now work directly with technology companies to package and sell AI tools, including hardware and software, to friendly nations.

Reorganizing Federal AI Priorities

One of the government's AI institutes was renamed and given new responsibilities. The shift reflects the administration's focus away from safety and toward growth. Michael Kratsios, who worked on technology policy in the first Trump administration, has returned to help shape the current AI strategy.

The export plan shows a major policy reversal. The previous administration had generally restricted the sale of advanced AI technology to other countries, concerned it could be misused. The Trump approach is to share AI widely with allied nations as a way to strengthen relationships and compete globally.

What Remains Unchanged

Not all federal AI rules have been eliminated. Some earlier orders that guide how the government uses AI internally are still in effect. These orders direct federal agencies to keep track of how they use AI and to build expertise in the technology.

The Department of Homeland Security completed a report on AI risks to critical infrastructure in response to the order that was cancelled. The agency will continue this safety work using older cybersecurity rules that remain in place.

Federal agencies can still collect information about how AI systems are being used and what results they produce. This information gathering, set up by earlier orders, helps the government understand both the benefits and risks of AI as officials plan for the future.

What This Means: The Bet Behind This Policy

The broader context here is that the Trump administration is betting on a different strategy than its predecessor. The previous approach said: "The government should set rules for AI development to prevent harm." The new approach says: "Competition and markets will naturally lead companies to develop AI responsibly, without the government needing to mandate safety practices."

This is not a new pattern in technology policy. When the internet became commercial in the 1990s, the Clinton administration took a similar approach—few rules, rapid growth. Years later, people faced problems—from privacy violations to the spread of misinformation—that might have been prevented with more oversight. The Trump administration believes AI is different; that American companies will innovate faster without federal roadblocks.

For AI companies, this change removes compliance burdens. They no longer need to conduct mandated safety tests or file required reports with the government. It also opens doors for international business. With government backing, American firms can more easily sell AI technology to other countries.

One thing worth noting is that this policy shift creates instability for the long term. Companies that invest billions in AI technology need to know what the rules will be five or ten years from now. When policies swing dramatically with each new administration, it becomes harder for businesses to plan. This short-term freedom could come with a long-term cost in terms of confidence and coordination.

The way the U.S. regulates AI will likely continue to change as administrations change. This volatility may ultimately make it harder for America to lead global conversations about responsible AI development—a challenge that will outlast any single administration as AI continues to reshape nearly every industry and sector worldwide.