What's Happening?
The Centers for Medicare and Medicaid Services (CMS) is re-evaluating its approach to Artificial Intelligence (AI) adoption, moving beyond simply tracking the number of employees using AI tools to measuring the actual outcomes and benefits derived from
these technologies. According to FedScoop, CMS Chief Information Officer Patrick Newbold highlighted this shift during a panel discussion in Bethesda, Maryland. The agency has reported that approximately 80 percent of its workforce interacts with AI tools, including its internal AI assistant, CMS Chat. Additionally, CMS has over 70 documented AI use cases in the Department of Health and Human Services’ 2025 AI use case inventory and utilizes generative AI tools through discounted contracts under the General Services Administration’s OneGov initiative. Newbold emphasized that while usage data remains valuable, it is no longer the primary metric for success. Instead, CMS is prioritizing the achievement of specific outcomes, viewing AI as a means to an end rather than an end in itself.
Why It's Important?
This strategic shift by CMS is significant as it underscores a maturing understanding of AI implementation within federal agencies. By focusing on outcomes, CMS aims to ensure that AI investments translate into tangible improvements for the populations it serves, rather than merely indicating technological adoption. This approach could set a precedent for other federal civilian agencies, encouraging a more results-oriented deployment of AI. The move also highlights the challenges associated with evaluating the true impact of AI, particularly in complex governmental operations. Furthermore, the impending expiration of discounted AI agreements under GSA’s OneGov initiative raises concerns about future costs and the need for robust governance practices to manage spending and ensure transparency in AI usage. This focus on outcomes will be crucial for justifying continued investment and demonstrating the value of AI in public service.
What's Next?
CMS plans to continue refining its outcome-focused approach to AI adoption. The agency is utilizing an internal initiative called Tech Pulse, launched approximately 20 months ago, to track technology-related outcomes. This performance management forum, inspired by a similar effort at the Social Security Administration, provides a structured method for monitoring and advancing key IT goals using objectives and key results as its measurement framework. As the discounted agreements for AI systems under OneGov are expected to expire, CMS will need to navigate future procurement challenges and potential cost increases. Newbold acknowledged the lack of clarity from AI model developers regarding long-term pricing and stressed the importance of maintaining strong governance practices to ensure clear visibility into AI usage and prevent unpredictable spending escalation. The agency is confident in its current infrastructure to manage this oversight effectively.
Beyond the Headlines
The shift from usage metrics to outcome-based evaluation in AI adoption by CMS reflects a broader trend in technology implementation, moving beyond mere deployment to demonstrating concrete value. This evolution has ethical and practical implications, as it forces agencies to clearly define the societal benefits and operational efficiencies they expect from AI, rather than adopting technology for its own sake. It also brings to the forefront the need for robust measurement frameworks and accountability mechanisms to ensure that AI serves public interest effectively and responsibly. The uncertainty surrounding future AI procurement costs, particularly with the expiration of discounted contracts, could also lead to a more cautious and strategic approach to AI investment across the federal government, emphasizing long-term value and sustainable deployment over rapid, unmeasured adoption. This could foster a more mature and responsible AI ecosystem within the public sector.
