Federal health agencies are rapidly adopting artificial intelligence, but their rush to deploy these tools is outpacing their willingness to govern them.

The Department of Health and Human Services is experiencing an unprecedented surge in AI adoption. At the Food and Drug Administration, documented AI use cases jumped 148% in fiscal year 2025. On paper, the federal health apparatus is moving at breakneck speed. Employees now have access to department-wide generative AI, alongside custom agency programs like the FDA’s Elsa 4.0 and the CDC’s ChatCDC.

But look closer at the data, and the narrative of rapid progress begins to fray.

The Compliance Illusion

A significant portion of this sudden spike might not represent new technology at all. Instead, analysts note it likely reflects stricter reporting mandates that forced agencies to finally disclose tools they were already using in the background.

More concerning is the massive gap in oversight. While agencies were quick to list their new AI projects, they largely left critical risk management and impact assessment fields blank in their official inventories.

They are deploying the technology first and delaying the safety guardrails.

The Governance Gap

This is not just a bureaucratic oversight. As federal health agencies pilot “agentic” AI capable of autonomous tasks, the lack of documented risk assessments becomes a major liability.

If the regulators of public health will not fill out their own safety checklists, they lose the moral authority to police the private sector’s use of the same technology. The rush to adopt has created a dangerous governance vacuum at the very top of healthcare oversight.