Federal AI: Selective Adoption

Federal agencies are narrowing their artificial intelligence plans to a smaller set of uses they can actually run, govern and measure. (federalnewsnetwork.com) Laura Stash wrote in Federal News Network on April 14 that agencies are moving from general adoption to “smart, intentional uses,” after building up talent through steps including the federal Tech Force and expanding agency use cases. (federalnewsnetwork.com) Mike Scopa, senior vice president of engineering at Deltek, wrote the same day in GovCon Wire that cloud computing is becoming essential for government contractors, but cost, compliance and legacy systems still block migration. (govconwire.com) Artificial intelligence systems need usable data and enough computing capacity to process it, and both commentaries describe old on-premises systems as a bottleneck for federal deployment. Federal News Network separately reported on April 1 that “data readiness” is the starting point for adoption and tied that work to the White House’s 2025 priorities for infrastructure and data centers. (federalnewsnetwork.com) The policy backdrop is already in place. The Office of Management and Budget’s March 2024 memo, M-24-10, required agencies to set up artificial intelligence governance, including naming a Chief Artificial Intelligence Officer within 60 days. (whitehouse.gov) The White House updated that approach on April 3, 2025, in memo M-25-21, which said agencies should cut bureaucratic bottlenecks, treat governance as an enabler of safe innovation, and set clear workforce expectations for appropriate artificial intelligence use. (whitehouse.gov) That makes the current argument less about buying one big artificial intelligence platform and more about choosing narrow jobs that fit existing missions, then upgrading the plumbing underneath. Federal News Network’s April 14 piece frames that as the practical path for agencies already under pressure to show results. (federalnewsnetwork.com) The standards side points in the same direction. The National Institute of Standards and Technology says its Artificial Intelligence Risk Management Framework is meant to help organizations manage risks tied to design, development, deployment and use, not to serve as a one-time procurement checklist. (nist.gov) For contractors, Scopa said the cloud question is also a market question: firms that cannot modernize face higher operating friction in a federal environment that increasingly expects secure, scalable digital delivery. (govconwire.com) The thread running through both commentaries is simple: in federal technology, artificial intelligence is starting to look less like a wholesale overhaul and more like targeted deployments built on modern infrastructure. (federalnewsnetwork.com)