HHS Publishes New AI Strategy for Expanding AI Adoption
Last week, the Department of Health and Human Services (HHS) published its artificial intelligence (AI) strategy – a plan for increasing AI adoption within the HHS to improve efficiency and cut costs. The AI plan will see AI tools shared across all HHS departments, including the CDC, CMS, FDA, and NIH, with the goal of “supercharging internal operations through an AI-empowered workforce”.
This approach, dubbed OneHHS, is intended to unify the HHS through shared AI infrastructure, streamline workflows, improve cybersecurity, and modernize the nation’s public health systems. While OneHHS has an initial internal focus, the HHS will seek to improve engagement with private sector stakeholders to develop new AI tools. The HHS strategy is based on five strategic pillars:
- Strengthening governance and risk management
- Developing infrastructure and platforms around users’ needs
- Promoting workforce development and burden reduction to improve efficiency
- Fostering health research and reproducibility through gold standard science
- Modernizing clinical and public health for better outcomes
The goal of the OneHHS approach is to standardize how federal agencies evaluate, adopt, and interface with private AI tools. “By establishing AI-enabling infrastructure, we are streamlining processes and reducing inefficiencies. In parallel, we are training our workforce to use AI at all levels, to update workflows, and automate tedious tasks,” said HHS Chief AI Officer, Clark Minor. “We will better deliver health care and human services through this AI-equipped workforce.”
In fiscal year 2024, the HHS had 271 active or planned implementations of AI, which are anticipated to increase by around 70% in fiscal year 2025. The HHS has made ChatGPT available for all HHS employees, the FDA has confirmed that it will soon adopt agentic AI tools to improve efficiency for pre-market reviews, and AI will be adopted by the CDC to accelerate the analysis of public health data.
“AI has the potential to revolutionize health care and human services, and HHS is leading that paradigm shift,” said Deputy Secretary of Health and Human Services Jim O’Neill. “By guiding innovation toward patient-focused outcomes, this Administration has the potential to deliver historic wins for the public—wins that lead to longer, healthier lives.”
Initially, the HHS will be conducting a preliminary assessment of the AI tools currently used within the HHS to assess the current level of HHS AI maturity. Under Pillar 1, the HHS has established a governance board that includes representatives from all HHS departments. The board will meet twice a year and will monitor the use of AI tools, establish risk management protocols for high-impact AI, and streamline approvals for new AI projects.
The main goal of pillar 2 is to create an HHS data infrastructure aligned to FAIR data principles. The infrastructure will centralize data and governance to improve efficiency and cut costs by eliminating duplication, and will provide AI-ready access to study data, claims, EHR, registries, surveillance, and human services data, with governance and monitoring embedded for cost effectiveness. Pillar 3 is concerned with promoting workforce development and burden reduction for efficiency, establishing role-based training pathways, the deployment of secure, approved AI tools that are integrated with HHS data, establishing service desks and libraries for common tasks and prompts, recruiting top AI talent, and developing tracking mechanisms to evaluate the quality and value of deployed AI tools.
Pillar 4 is concerned with fostering health research and reproducibility through gold standard science, and aims to set a global benchmark for the use of AI in biomedical research and development. The gold standard principles will be used across all areas of AI-augmented research. Under pillar 5, the HHS plans to increase the use of AI to deliver measurable improvements in population health and individual patient outcomes, making AI tools available to support clinicians and aid decision-making, and improve efficiency.
“Artificial intelligence has enormous potential to strengthen our nation’s health-care system,” said Rep. Jay Obernolte (R-CA), Chairman of the Bipartisan House Task Force on Artificial Intelligence. “I’m encouraged that HHS is taking a thoughtful, outcomes-driven approach that prioritizes transparency, rigorous risk management, public trust, and respect for Americans’ health data.”
