Public health departments across the United States are set to trial generative AI tools under a new initiative called the Public Health Use Case and Learning Scaling Engine, or PULSE. Led by the Coalition for Health AI (CHAI), the program involves OpenAI, Anthropic, and Accenture, and will run pilots in 10 state, local, tribal, or territorial jurisdictions. The goal is to produce implementation guidance for public health agencies considering similar AI deployments.
What Happened: PULSE Program Details
OpenAI and Anthropic have donated 10 enterprise licenses each, with capacity for up to 2,000 public health practitioners. Accenture will oversee participant onboarding and help develop playbooks based on the trial results. The pilots will focus on five use cases: biosurveillance and drug-wave prediction, social determinants of health (SDoH) mapping, operations efficiency and community-feedback analysis, public communications and multilingual translation, and automated clinical-data retrieval using a FHIR query engine. However, CHAI has not specified which products, model versions, or configurations will be used, nor how the two providers will be assigned across the pilots.
Why It Matters: Trust, Governance, and Execution
This program is significant because it directly addresses the challenges of trust, governance, and execution that are critical for AI adoption in the public sector. Dr. David Lakey, a former Texas health commissioner, noted, “Every major technological transformation succeeds or fails based on trust, governance and execution. PULSE will support agencies in this endeavour, and is specifically designed for practical implementation.” Yet, many governance details remain undefined. CHAI has not published separate evaluation, privacy, security, or human-review requirements for the five use cases. The application of HIPAA will vary depending on the agency, data involved, and function performed, adding complexity to compliance.
Our Interpretation: Data Governance and Security Are Key Variables
XPLAIN AI sees the opacity around data handling as the most critical point in this announcement. It is unclear whether the pilots will use identifiable records, de-identified information, synthetic data, or aggregated datasets. Retention periods, access controls, audit arrangements, data-storage requirements, and rules for submitting protected health information have not been defined. While OpenAI and Anthropic state that inputs and outputs from their enterprise services are not used for model training by default, how these policies will be configured in the PULSE deployments remains unspecified. This ambiguity could lead to compliance risks in real-world operations. Consequently, demand for data security and governance solutions may rise, benefiting vendors in that space. Additionally, the pilot results will shape the competitive landscape between OpenAI and Anthropic in the public health AI market.
Beneficiaries and Risks: Who Gains and Who Faces Challenges
Direct beneficiaries of the program are OpenAI and Anthropic, which gain real-world use cases and product validation in the public health sector. Accenture (ACN) also stands to benefit from its consulting and systems integration role. On the risk side, traditional healthcare IT vendors such as Oracle (ORCL) and privately held Epic Systems may face disruption if generative AI approaches replace existing workflows. Conversely, data security and governance firms like Palo Alto Networks (PANW) and CrowdStrike (CRWD) could see increased demand as public health agencies tighten security requirements.
Counter Scenario and Uncertainty: Governance Gaps Could Derail Progress
The program’s success is far from guaranteed. CHAI has not yet published evaluation metrics or success criteria for each use case. Pilots are scheduled to begin in autumn 2026, with playbooks expected in 2027—but without clear governance and security requirements, participating agencies may struggle to implement findings. The ambiguity around HIPAA coverage also introduces legal risk. If pilot results disappoint or a data breach occurs, public-sector AI adoption could slow significantly. Moreover, the lack of transparency around model selection and data handling may erode trust among stakeholders.
Key Indicators to Watch
- List of participating jurisdictions: Which states or local governments join will indicate the program’s scale and representativeness.
- Evaluation criteria for each use case: Whether CHAI publishes specific metrics will be crucial for credibility.
- Data protection and privacy policies: Clear guidelines on HIPAA applicability and data handling are essential.
- Model versions and configurations: The specific models used will affect performance and risk profiles.
#PublicHealthAI #OpenAI #Anthropic #GenerativeAI #HealthcareIT #DataGovernance #HIPAA #AIPolicy
Sources
- US public health agencies to test OpenAI and Anthropic AI models — AI News · News coverage · Mon, 20 Jul 2026 10:00:00 +0000
Written by: XPLAIN AI Editorial Team · Reviewed by: XPLAIN AI Editorial Desk
This content was drafted with AI assistance based on publicly available sources and reviewed under XPLAIN AI's editorial standards.