The White House of the United States issued two new memoranda on April 3, 2025, providing guidance to federal agencies on the responsible use and procurement of artificial intelligence (AI), replacing the previous guidance of the Biden administration while maintaining its fundamental structures. The new guidelines, designated M-25-21 and M-25-22, follow Donald Trump's January executive order, which set out to develop a federal AI "action plan" focusing on three main priorities: innovation, governance, and public trust.
The new guidelines retain key elements established by the Biden administration, including the roles of Chief AI Officers (CAIOs). However, their responsibilities have been transformed to serve as change agents and AI advocates, not overseers of layers of bureaucracy—as stated in the White House briefing on April 7. The procurement directive (M-25-22) requires that within 100 days, the General Services Administration (GSA), in collaboration with the OMB, develop a plan to issue publicly available guidance to support the procurement workforce in acquiring AI systems. It sets a 200-day deadline for creating an online repository to facilitate sharing AI procurement information and tools.
The guidelines include risk management requirements for "high-impact AI" applications, which may have a significant impact upon deployment, and, similar to the Biden administration's guidance, deem inappropriate use to be discontinued. However, the new directive gives agencies one year to document minimum practices. Quinn Anex-Ries, a senior policy analyst at the Center for Democracy and Technology, commented: while it is encouraging that this updated guidance requires agencies to address risks when deploying AI systems proactively, recent reports indicate that DOGE's use of AI may not meet these basic transparency and risk mitigation practices—expressing concern about the practical implementation of the guidelines.
Sources:
1.
malder
2.
By: Grace Dille
3.
Alexandra Kelley
