The White House has issued sweeping new guidelines to accelerate the federal government’s adoption of AI. The directives, issued by the Office of Management and Budget (OMB) on April 3, provide a roadmap for implementing President Donald Trump’s January executive order aimed at removing barriers to AI deployment and boosting U.S. dominance in the field.

The memo, titled M-25-21: Accelerating Federal Use of AI through Innovation, Governance, and Public Trust, lays out deadlines and policies for federal agencies to not only adopt AI tools that improve public services but to do so while maintaining strict oversight on privacy, civil rights, and national security.

“Agencies must lean forward on adopting effective, mission-enabling AI,” the memo states, emphasizing a need to cut bureaucratic red tape and tap into American innovation.‍

Three Pillars: Innovation, Governance, and Public Trust

The guidelines are built around three pillars: innovation, governance, and public trust. Agencies are instructed to eliminate unnecessary internal restrictions that delay or limit the use of AI in mission-critical operations. They must also designate Chief AI Officers to lead adoption efforts and form internal governance boards to oversee AI use across departments.

Under the new framework, AI is not just a back-end tool—it’s central to the future of public service delivery. The OMB urges agencies to prioritize AI projects that reduce costs, streamline processes, and enhance citizen services. Importantly, the guidelines also push agencies to use AI developed and produced in the United States, in line with the administration’s “Buy American” policy.‍

AI Strategies, Inventories, and Risk Management

Every agency covered by the Chief Financial Officers Act is required to publish a public AI strategy within 180 days. These strategies must identify high-impact use cases, current AI maturity levels, and plans for workforce development and infrastructure upgrades.

The guidelines further require agencies to maintain a publicly available AI use case inventory, and to implement strict risk management practices for what the government labels “high-impact AI”—defined as AI systems that materially affect civil rights, public safety, or access to government services.

If such AI systems fall short of compliance, the memo mandates their suspension until proper safeguards are restored.‍

Generative AI, Transparency, and Ethics

In a nod to the rapid rise of generative AI technologies, the memo gives agencies 270 days to develop clear policies on how such tools may or may not be used. It also calls for agencies to conduct AI impact assessments, establish human oversight mechanisms, and create pathways for individuals to appeal decisions made—or influenced—by AI.

To build public trust, agencies must solicit feedback from users and the broader public. The memo encourages transparency through usability testing, public hearings, and online comment periods.‍

The Broader Picture

The new policy framework follows President Trump’s Executive Order 14179, Removing Barriers to American Leadership in Artificial Intelligence, signed in January. Together, these actions underscore the administration’s ambition to assert U.S. leadership in a field widely seen as a strategic cornerstone for global competitiveness, economic growth, and military innovation.

AI-driven transformation of federal services has already been underway for years, but the latest directive accelerates both pace and scale. With rising global competition and concerns about China’s advancements in AI, the US is betting on public sector modernization as a key pillar of technological dominance.