Optimizing Public Health Strategies for a Healthier America

Public health departments throughout the United States are set to participate in a new initiative that will test generative AI tools. This program is a collaboration between the Coalition for Health AI, OpenAI, Anthropic, and Accenture.

Known as the Public Health Use Case and Learning Scaling Engine (PULSE), this initiative will facilitate trials in 10 state, local, tribal, or territorial jurisdictions. The primary goal of the program is to provide guidance for public health agencies looking to implement similar AI deployments.

OpenAI and Anthropic have generously donated 10 enterprise licenses with the capacity to accommodate up to 2,000 public health practitioners. Accenture will be responsible for overseeing participant onboarding and assisting in the development of playbooks based on the trial results.

The program will grant public health practitioners access to enterprise AI products from OpenAI and Anthropic. However, the specific products, model versions, and configurations to be used have not been disclosed by CHAI. It remains unclear how these providers will be assigned across the pilot programs.

Dr. David Lakey, a former Texas health commissioner, emphasized the importance of trust, governance, and execution in technological transformations. He believes that PULSE will provide valuable support to agencies, offering practical implementation strategies.

The Coalition for Health AI’s leadership council will select the participating jurisdictions and assign practitioners to communities focused on five specific use cases. These include biosurveillance and drug-wave prediction, social determinants of health mapping, operations and efficiency analysis, public communications and multilingual translation, and automated clinical data retrieval and FHIR query engine.

The National Institute of Standards and Technology (NIST) recommends evaluating AI systems based on their intended use, operating environment, affected parties, and potential consequences. CHAI has not yet outlined the evaluation, privacy, security, or human-review requirements for the five PULSE use cases.

FHIR, an HL7 standard, is used for exchanging healthcare information electronically between compatible systems. PULSE incorporates an automated clinical data retrieval and FHIR query engine, but the specifics of generative AI’s role in this workflow have not been defined.

It is essential to note that HIPAA requirements must be adhered to when handling electronic health information. However, the application of HIPAA regulations will vary depending on the participating organization and the specific workflow being addressed.

Felipe Millon, OpenAI’s head of government go-to-market, expressed the belief that AI should be useful, safe, and accessible for addressing society’s most critical challenges. The licenses donated by OpenAI aim to assist public health organizations in evaluating AI tools through a structured process.

The pilots for the PULSE program are scheduled to commence in autumn 2026. CHAI anticipates releasing playbooks based on the trial outcomes in 2027, which will serve as reference materials for other public health agencies.

Governance and evaluation criteria for the pilots have not been published by CHAI. There is a lack of clarity regarding how model outputs will be reviewed and whether human oversight will be required for certain functions.

Elizabeth Kelly, Anthropic’s head of beneficial deployments, highlighted the importance of introducing AI into public health workflows with care and proper guardrails. She believes that PULSE will allow practitioners to test AI tools in their environments with privacy, governance, and responsible-use measures incorporated from the outset.

Data from the National Association of County and City Health Officials indicates that nearly 40% of local health departments are not currently using AI. PULSE aims to provide support in the form of enterprise licenses, onboarding assistance, peer communities, and implementation playbooks.

The initiative targets various public health agencies, including state and territorial health departments, county and municipal agencies, tribal authorities, Indian health organizations, and large city health departments. The ultimate goal of PULSE is to convert findings from 10 jurisdictions into guidance for broader use.

The Coalition for Health AI is actively working on governance standards for healthcare AI, with plans to develop guidance covering eight governance areas through workshops and working groups. The organization has begun publishing playbooks on organizational AI policies, governance structures, and internal resources.

In conclusion, the PULSE program represents a significant step towards integrating AI tools into public health workflows. By providing practical experience with AI technologies, the initiative aims to help agencies navigate the challenges and opportunities presented by AI in healthcare.