The U.S. Department of Health and Human Services grew its artificial intelligence use cases by up to 148 percent between fiscal years 2024 and 2025, yet agencies left critical risk management and oversight fields largely blank in federal inventories. This gap raises direct accountability concerns as high-stakes systems, including automated Medicaid eligibility platforms and drug safety monitors, move into active operation.
Accelerated deployment across agencies
The Food and Drug Administration saw the steepest growth in AI adoption at 148 percent, followed by the Centers for Disease Control and Prevention at 87 percent. The Centers for Medicare and Medicaid Services increased AI applications by 78 percent, while the National Institutes of Health grew by 51 percent. This acceleration followed a 2025 executive order directing the Office of Management and Budget to issue guidance on federal AI deployment.
Deployment maturity varies significantly among these bodies. The FDA successfully deployed 63 percent of its AI portfolio. In contrast, 55 percent of CMS systems remain in pre-deployment stages. The NIH retired 16 percent of its historical AI use cases, demonstrating active lifecycle management as older systems become obsolete.
Shift toward autonomous systems
Generative AI and natural language processing dominate agency portfolios, primarily for drafting documents and automating communication. A newer category, agentic AI capable of taking autonomous actions, has emerged across the CDC, CMS, and NIH. This signals a shift from tools that merely assist human decision-making to systems that can execute tasks independently.
CMS has positioned itself as an "AI first" organization in its strategic framework through 2031, committing to train thousands of employees on AI technologies. Meanwhile, commercial AI-driven cybersecurity tools represent the most widely deployed category across HHS. Platforms including CrowdStrike Falcon and Microsoft Defender currently support an estimated 50,000 or more users.
The governance gap
The fiscal year 2025 inventory introduced new fields to track risk management, including impact assessments, independent reviews, ongoing performance monitoring, and appeal processes. Agencies left most of these fields blank. The inventory also removed procurement tracking, making it harder to connect AI systems to specific federal contracts and spending.
The inventory requires a "high-impact" designation for systems that could significantly affect rights or safety. Of 447 total use cases, only two received this designation, including the CMS iVeri-Fi automated identity verification platform used in Medicaid eligibility. Another 13 systems were presumed high-impact but justified otherwise, such as FDA AI-driven video analytics for law enforcement, which officials deemed lower risk because human reviewers verify the outputs.
Federal policy continues to advance AI for Government integration alongside tightening security requirements. In the first half of 2026, an executive order gave agencies 30 to 60 days to strengthen government systems' AI-enabled cybersecurity. This mandate applies broadly to all healthcare companies, including rural hospitals and critical infrastructure providers.
Why this matters for healthcare professionals
Healthcare operators and administrators must prepare for the downstream effects of these federal deployments. As federal agencies embed AI for Healthcare into workflows governing drug approvals, disease surveillance, and Medicaid eligibility, the lack of mature oversight frameworks increases operational and compliance risks. Organizations should audit their own vendor contracts and internal AI policies now, ensuring human-in-the-loop verification remains standard practice before automated systems face regulatory scrutiny.
Your membership also unlocks:
