The White House is currently working on an executive order that seeks to transform how artificial intelligence companies release their advanced models. One of the primary stipulations of this order is that developers will have to submit their AI models to the federal government at least 90 days prior to public launch.
This forthcoming executive order is focused on the intersection of AI innovation and national security, particularly addressing the cybersecurity threats that powerful AI models may pose. The growing capabilities of AI raise valid concerns about their potential misuse in offensive cyber activities. These could range from automating the discovery of system vulnerabilities to executing widespread phishing attacks.
While the main provision obligates pre-release sharing, the overall framework is characterized more as a voluntary approach rather than a strict mandate. This advance notice is designed to allow federal agencies sufficient time to evaluate any potential national security risks inherent in new AI models before they become widely accessible.
As discussions around the order continue, it's important to note that the final iteration might differ from what has been proposed so far.
#What are the implications of this order on AI governance?
This executive order will not be the first action concerning AI governance under the current administration. In January 2025, a previous executive order aimed at enhancing the United States' leadership in artificial intelligence was signed. This earlier directive took a significantly different approach compared to the AI policies under the Biden administration, opting to rescind certain restrictions imposed by the previous government that were viewed as hindering innovation.
A second cybersecurity order came into effect in June 2025, emphasizing safe software development practices and the role of AI in cybersecurity.
The upcoming order introduces a new element to the existing framework. While the January order focused on elevating the nation’s posture in AI and the June order concentrated on defense strategies, this new directive emphasizes oversight. It marks the first instance of requiring developer engagement with government agencies prior to the launch of AI models.
#How might this affect investors in AI?
It's essential to highlight that the executive order does not specify any particular cryptocurrencies or blockchain technologies. However, the introduction of a 90-day pre-release sharing requirement, even on a voluntary basis, could disrupt the speed at which AI products enter the market. For publicly traded AI firms, this could lead to postponed product introductions, increased compliance expenses, and unpredictable outcomes regarding potential government evaluations during the 90-day review period.
Moreover, decentralized AI projects, which are gaining traction across various blockchain networks, function in a fundamentally different manner than traditional centralized AI companies. The implications of a voluntary sharing protocol for open-source or decentralized AI development remain uncertain, especially since there may not be a single corporate entity overseeing the review process for such models. Such uncertainties should be closely monitored as the final details of the order emerge.