The Office of Management and Budget has issued two memos to facilitate the responsible adoption of artificial intelligence tools and drive efficient AI acquisition in the government.
In a fact sheet published Thursday, the White House said the policies mark the executive branch’s transition to a pro-innovation, pro-competition approach while mitigating unnecessary bureaucratic restrictions on the use of AI and enhancing U.S. superiority in AI innovation.
Table of Contents
Fostering Responsible AI Adoption
M-25-21 provides agencies with the tools needed to advance AI innovation while maintaining strong protections for the civil rights, privacy and civil liberties of Americans.
Under the new memo, chief AI officers should promote agencywide AI adoption and innovation for lower risk AI; advise on AI investments and spending; and mitigate risks for higher-impact AI.
The policy directs agencies to produce an AI adoption maturity assessment to monitor progress and needs. It also introduces a “high-impact AI” category to track AI use cases that demand heightened due diligence due to potential impacts on the safety or rights of citizens.
Advancing Efficient AI Acquisition
M-25-22 offers guidance on how agencies could improve their ability to acquire AI tools rapidly, competitively and responsibly.
Under the memo, agencies must support a competitive AI marketplace by maximizing the use of American-made AI systems and services. They should also use performance-based acquisition techniques and establish an online shared repository of tools and resources to help with AI procurement.
The policy outlines specific requirements that avoid vendor lock-in and seeks to eliminate burdensome agency reporting requirements to facilitate the acquisition process.
Related Articles
A U.S. government slide presentation shows that the Trump administration’s Golden Dome next-generation missile defense shield will include four layers: one satellite-based platform and three land-based missile systems, Reuters reported Tuesday. The missile defense shield is expected to cost $175 billion and be completed by 2028. According to the slides, one of the layers will be a space-based sensing and targeting layer for missile warning and tracking, and missile defense. Meanwhile, the three ground-based layers consist of radar arrays, missile interceptors and lasers. The slides presented to 3,000 defense contractors in Huntsville, Alabama, show 11 short-range missile batteries located across
The Information Technology Industry Council, or ITI, has provided the Trump administration and newly confirmed White House National Cyber Director Sean Cairncross with several policy recommendations to strengthen U.S. cyber defenses. 4 Strategic Priorities to Help ONCD Address Cybersecurity Needs ITI said Tuesday it organized its recommendations in four sections to help the Office of the National Cyber Director, or ONCD, address cybersecurity needs: lead with strength and speak with one voice; cut red tape and secure the nation; leverage public-private collaboration as a strategic asset; and defend against real and emerging threats. “ITI’s new recommendations provide a results-driven action
A new OneGov agreement between the General Services Administration and Anthropic will provide the federal civilian executive, legislative and judiciary government branches with access to Claude for Enterprise and Claude for Government for only a dollar. Under the agreement, all eligible participating branches will gain up to a year of access to the Anthropic offerings for a nominal fee, including the company’s frontier models with continuous updates as new capabilities are released. The company will also provide agencies with technical support to help them implement artificial intelligence into their productivity and mission workflows, GSA said Tuesday. Accelerating Government AI Adoption