Yesterday, President Trump signed an executive order establishing a voluntary framework under which technology companies can provide the federal government with access to advanced artificial intelligence (AI) models up to thirty days before those systems are released to the public. The stated goal is to improve national security and cybersecurity by allowing government agencies to evaluate the potential risks posed by increasingly powerful AI technologies.

Under the new framework, federal agencies, including the Department of Defense and National Security Agency, will review participating companies’ AI models and assess whether they present security vulnerabilities or could be misused by hostile actors. 

The executive order represents a change from the Trump Administration’s past approach to regulating AI. Prior to this order, the administration had taken a more relaxed approach to monitoring and regulating new AI technologies.

While this executive order only asks for tech companies’ cooperation, the underlying message appears to be those companies who don’t cooperate could face additional scrutiny from the federal government…

The SpaceX IPO is days away... Space stocks are ripping... And the NASA ETF is up nearly 70% in eight weeks. But the man behind the most profitable research in Stansberry Research history says most investors are about to make a serious mistake. On June 4, he's breaking nearly four years of public silence to share what he calls the most important buy call of his career, and a make-or-break issue for Elon Musk that no one is talking about. Learn more here.

When President Trump returned to office in 2025, his administration emphasized reducing regulations that it believed could slow innovation. One of Trump’s first AI-related actions was an executive order titled ‘Removing Barriers to American Leadership in Artificial Intelligence,’ which sought to eliminate what the administration viewed as unnecessary restrictions on AI development and deployment. The White House argued that excessive regulation could weaken America’s competitive position, particularly in relation to China.

The scale and scope of AI growth in just the past year probably prompted the change in approach. While AI has been gathering momentum since the first appearance of large language models in 2022, the past year has seen not just significant advances in AI capabilities but also an enormous increase in the amount of AI infrastructure being built by technology companies. Technology is always ahead of the law, which is a challenge for the government because lawmakers and regulators often do not fully understand the implications of new technologies such as AI.

AI’s improved capabilities have led government officials to worry about the potential impact on national security. As AI systems became more capable, policymakers increasingly worried that they could be used to discover cyber vulnerabilities, automate sophisticated attacks, or provide strategic advantages to foreign adversaries. The executive order reflects growing concern within the U.S. government that advanced AI development requires at least some level of coordination between industry and national security agencies.

Technology companies may be reluctant to share new AI technology with the government due to concerns about competition. If their competitors find out about a new AI program before it’s launched, they could copy it or try to interfere with the roll-out. 

Technology companies themselves appear to share at least some of the government’s concerns. Anthropic publicly stated its AI platform Mythos was too dangerous for open release. As the competition for AI dominance intensifies, particularly between the United States and China, this executive order is likely to shape the debate over how much government oversight advanced AI systems should face in the years ahead.