US public health agencies will pilot generative AI systems from OpenAI and Anthropic through a new program called PULSE, the Public Health Use Case and Learning Scaling Engine. The initiative involves the Coalition for Health AI, both AI companies, and consulting firm Accenture, deploying tools across 10 state, local, tribal, and territorial health jurisdictions.
PULSE represents an early effort to move AI beyond tech companies into government healthcare operations. Public health departments handle disease surveillance, emergency response, epidemiology, and health equity analysis. Large language models could theoretically accelerate document processing, help analyze disease patterns, or improve communication during outbreaks. The trials will test whether these systems actually deliver value in real health operations or simply create compliance overhead.
The program structure matters. Ten jurisdictions is small enough to be manageable but large enough to surface real implementation problems. The inclusion of tribal and territorial health systems signals an attempt to reach underserved areas, though scaling across fragmented public health infrastructure remains difficult. Many state and local agencies run decades-old systems with limited IT staff.
Accenture's involvement suggests this leans toward consulting and integration services rather than pure technology transfer. The company often sells implementation expertise alongside software, which could either smooth deployment or add unnecessary complexity.
Key questions remain unanswered. The program details are sparse, but typical public health pilots examine whether AI improves response times, reduces administrative burden, or strengthens equity in care. Privacy and data security will be critical. Health data is protected under HIPAA, meaning any AI system needs careful handling of patient information, data retention policies, and audit trails. Both OpenAI and Anthropic have published safety practices, but applying them to sensitive government health data requires specific configurations and oversight.
Results will likely take 12-18 months to materialize. Public health agencies move deliberately on new tech, and rightfully so. The