The Rise of AI in Federal Health Agencies: A Closer Look at Elsa and Beyond
In June 2025, the U.S. Food and Drug Administration (FDA) made a significant leap into the realm of artificial intelligence (AI) with the launch of "Elsa," a generative AI tool aimed at enhancing productivity and efficiency within the agency. This initiative marks a pivotal moment as it represents the FDA’s first internal deployment of a commercial large language model. However, the story of AI within federal agencies, particularly the U.S. Department of Health and Human Services (HHS), stretches back much further. Since 1996, HHS has steadily integrated AI technologies into its operations, with a noticeable acceleration in adoption seen in recent years.
Historical Context of AI Adoption in HHS
The journey of AI utilization in HHS gained momentum with the issuance of a 2020 executive order that mandated federal agencies to catalog and publicly disclose non-classified uses of AI. This initiative was part of a broader push to embrace innovative technologies within government workflows, and it has borne fruit in the form of the "HHS AI Use Case Inventory." Published for the first time in 2022, this inventory has become an essential tool for tracking how various HHS sub-agencies are leveraging AI.
Health Secretary Robert F. Kennedy Jr. has been a staunch advocate for AI implementation since taking office, emphasizing its transformative potential in both internal and public-facing functions. His leadership has been instrumental in promoting a culture of innovation within the agency.
Surge in AI Use Cases
According to a recent report from the U.S. Government Accountability Office (GAO), HHS has experienced the most significant increase in generative AI applications among federal agencies. The number of AI use cases skyrocketed from 163 in 2023 to 270 in 2024, reflecting a nearly fourfold rise in just two years. This exodus into AI was largely spurred by the advent of generative AI tools such as OpenAI’s ChatGPT, which have demonstrated remarkable capabilities in language processing and task automation.
In April 2025, the White House’s Office of Management and Budget (OMB) issued two memoranda promoting strategic AI governance. These directives established guidelines for agencies to monitor AI use and appoint Chief AI Officers, thereby reinforcing transparency and ushering in a new era of technological oversight.
Deep Dive: How HHS Uses AI
HHS has categorized its AI applications into six distinct areas, each serving unique functions across various agencies. This systematic approach facilitates clearer understanding and implementation of AI technologies throughout the department.
1. Centers for Disease Control and Prevention (CDC)
The CDC employs AI mainly to enhance the effectiveness of its data analysis and outbreak prediction efforts. Notably, in 2023, the agency introduced "ChatCDC," an internal generative AI chatbot powered by OpenAI’s language models, designed to assist staff in operational tasks such as document summarization and literature reviews. Following the success of ChatCDC, a more sophisticated version integrating secure access to internal data is being piloted.
2. Centers for Medicare & Medicaid Services (CMS)
CMS applies AI technologies across a spectrum of activities, from processing customer complaints to identifying potential fraud. The agency has recently implemented machine learning algorithms to detect unusual billing patterns and assess Medicare claims for errors. Strikingly, in June 2025, collaboration between the Department of Justice and CMS led to the prevention of over $4 billion in fraudulent claims.
3. Food and Drug Administration (FDA)
As a crucial player in the regulatory landscape, the FDA has utilized AI to enhance its review processes for new drugs and medical devices. Elsa, the newest generative AI tool, is designed to assist staff in reading and summarizing internal documents. Notably, Elsa operates in a secure GovCloud environment, ensuring that sensitive information remains protected. FDA Commissioner Marty Makary has highlighted Elsa’s potential to expedite clinical review processes, although concerns have surfaced regarding the tool’s influence on regulatory decision-making.
Regulatory Oversight and Future Directions
The FDA not only aims to optimize its internal use of AI but is also establishing two cross-agency AI councils. One council will concentrate on regulatory strategies for AI-enabled products, while the other will guide the FDA’s internal AI applications. This proactive governance model aims to promote responsible AI use and maintain a balance between innovation and public safety.
The extensive tracking of AI applications throughout HHS serves a dual purpose: it provides transparency to the public while equipping policymakers with insights into how federal agencies leverage AI technologies. Collecting performance metrics and understanding the implications of these tools on productivity and outcomes will be critical for ongoing refinement and implementation of AI initiatives.
A Broader Impact on AI Use in Health Care
The rapid integration of AI within federal health agencies mirrors trends observed in the private sector, illustrating a wider shift toward digital transformation in service delivery. The ongoing evaluative processes, bolstered by executive mandates, reflect a commitment to ethical standards and effective use of technology in a fast-evolving landscape.
The articles and reports from organizations like the Bipartisan Policy Center (BPC) are invaluable resources for anyone keen to explore the impact of AI in health care and the broader policy frameworks that govern its use. For readers interested in contemporary developments and practical applications of AI, numerous online resources are available, including blogs, videos, and fact sheets to deepen their understanding.
As AI technologies continue to evolve, so do the opportunities and challenges they present in federal health agencies. Tracking these developments offers a glimpse into how AI may shape the future of health care, public policy, and potentially improve citizen engagement in the coming years.