What happened to the executive order for AI models? The White House was on the verge of signing a new executive order to establish a voluntary review for advanced AI models but ultimately decided against it.
On May 20, reports emerged about a draft from the administration that proposed a framework requiring developers of certain high-level AI technologies to submit their systems for review by federal agencies, including the NSA, up to 90 days before they were made public. The signing ceremony was scheduled for May 21, but it never took place.
The draft laid out a voluntary review system, meaning that while developers were encouraged to allow federal agencies to examine their advanced systems, they were not obligated to do so. The 90-day window was primarily designed to give agencies time to evaluate potential cybersecurity threats and assess national security ramifications.
Importantly, the draft included provisions that would prevent mandatory licensing or preclearance. Thus, companies could launch their AI products without requiring government permission, although an early look from regulators was advised for the most advanced systems.
Why was the executive order withdrawn? Just a few hours before the planned signing, David Sacks, who previously served as Trump's AI and crypto czar, alerted the President to concerns raised by industry leaders. These leaders argued that even a voluntary review process might slow down American AI development at a critical time when competition with countries, specifically China, is intensifying. Due to the rising apprehension about possible hindrances to US competitiveness, President Trump ultimately opted to halt the process.
The larger context of regulatory efforts shows that the administration had already attempted to ease regulations affecting US leadership in AI with an executive order issued on January 23, 2025. This aimed to counteract state-level regulations that could splinter a national approach to AI governance. The shelved executive order would have marked the first real attempt to impose a structured oversight mechanism yet. While there are valid concerns regarding national security about releasing powerful AI systems without regulatory oversight, the technology sector argues that any additional friction in the development could give an edge to competitors with fewer restrictions.
What are the implications for investors? The proposed voluntary review framework was already minimal governmental oversight, and if even that couldn’t progress due to lobbying, it raises concerns about the future of more stringent regulations. Presently, AI companies are functioning in a largely regulatory vacuum at the federal level.
Furthermore, there is a competitive angle to consider. The reasons given for pulling back on the executive order stressed that it would impede the ability of US firms to compete against their Chinese counterparts. However, it must also be noted that organizations and countries looking to purchase AI systems increasingly demand assurances of safety and reliability. A voluntary review framework, even if optional, could have served as a competitive edge rather than a limitation.