TechNewsReel
Live

CMS shifts AI strategy from usage rates to measurable mission outcomes

The agency is abandoning 'vanity metrics' to ensure AI investments deliver concrete results as discounted government deals expire.

TechNewsReel Newsroom · September 5, 2026

The Centers for Medicare & Medicaid Services (CMS) is shifting its artificial intelligence strategy to prioritize measurable outcomes over simple adoption rates. CMS CIO Patrick Newbold announced that the agency will no longer lead with the technology itself, but rather with the specific goals the technology is intended to achieve.

To manage this transition, CMS has implemented a performance forum known as "Tech Pulse." This framework utilizes Objectives and Key Results (OKRs) to monitor how AI serves as an enabler for agency goals. The shift comes at a time of high adoption; approximately 80% of the CMS workforce currently utilizes AI tools, including the agency's internal "CMS Chat" chatbot.

The push for governance

This strategic pivot follows a period of rapid expansion. CMS accelerated its AI integration through the GSA's OneGov initiative, which provided the agency with discounted access to foundational models. This early access allowed the agency to scale quickly, resulting in more than 70 AI use cases reported in the HHS 2025 inventory.

However, the financial landscape is changing. As the discounted OneGov deals expire, CMS faces significant uncertainty regarding long-term operational costs. Agency leadership is now implementing stricter governance measures to prevent costs from spiraling as the tools move from subsidized pilots to permanent infrastructure.

Why the shift matters

For the federal government, this move represents a critical maturation phase in AI adoption. By abandoning "vanity metrics"—such as how many employees have an account or how often a tool is opened—CMS is attempting to create a blueprint for justifiable spending. Linking AI usage directly to service improvements and mission success allows the agency to defend the high cost of these tools through population-level improvements.

As Patrick Newbold noted, the goal is to lead with outcomes first, viewing AI as a supporting tool rather than the primary objective. This approach ensures that technology is not deployed for its own sake, but only where it provides a tangible benefit to the agency's operations.

What's next

Industry observers will be watching how CMS defines its specific "outcome metrics" within the Tech Pulse forum. The agency's ability to quantify the exact value of its 70+ use cases will likely determine the scale of its future AI budget. While the agency has successfully integrated AI into the daily workflow of the vast majority of its staff, the next challenge lies in proving that this widespread usage translates into a more efficient healthcare system.

Sources

Get a notification when a big story breaks. A few a day at most — no spam.