**Public Health Agencies to Test Generative AI in New PULSE Programme**
A significant development in the application of artificial intelligence within the public sector has emerged, as public health departments across the United States are set to trial generative AI tools through a new initiative called PULSE. This programme, a collaboration between the Coalition for Health AI (CHAI), major AI developers OpenAI and Anthropic, and consulting firm Accenture, marks a substantial step in integrating advanced AI into national health infrastructure.
### **What is the PULSE Programme?**
PULSE stands for Public Health Use Case and Learning Scaling Engine. Its primary function is to support a series of trials taking place in 10 state, local, tribal, or territorial jurisdictions. The overarching goal is to create implementation guidance for public health agencies nationwide who are considering similar AI deployments. OpenAI and Anthropic have contributed 10 enterprise licences, capable of supporting up to 2,000 public health practitioners. Accenture will be responsible for onboarding participants and developing playbooks based on the trial results.
The programme will grant public health practitioners access to enterprise-grade AI products from the two leading AI providers. However, specific products and model versions have not yet been disclosed, nor has the exact allocation of providers across the different pilot projects been explained.
### **Five Key Public Health Use Cases**
The Coalition for Health AI’s leadership council will select the participating jurisdictions and assign them to one of five core use cases:
1. **Biosurveillance and Drug-Wave Prediction:** Using AI to monitor and predict the spread of diseases and drug trends.
2. **Social Determinants of Health (SDoH) Mapping:** Analysing non-medical factors like socioeconomic status and environment that impact health outcomes.
3. **Operations and Efficiency & Community-Feedback Analysis:** Streamlining administrative tasks and analysing community input to improve services.
4. **Public Communications and a Multilingual Translation Hub:** Enhancing public messaging and breaking down language barriers.
5. **Automated Clinical-Data Retrieval and a FHIR Query Engine:** Automating the extraction and querying of electronic health records using Fast Healthcare Interoperability Resources (FHIR) standards.
### **Governance, Evaluation, and Key Concerns**
The PULSE programme is scheduled to begin in autumn 2026, with playbooks expected to be released in 2027. However, the announcement has raised several important questions, as governance and evaluation metrics remain undefined. The programme does not yet detail how model outputs will be reviewed. For instance, it is unclear whether AI-generated public communications, translations, or clinical data will require mandatory staff verification before use.
While the NIST AI Risk Management Framework recommends evaluating AI systems by their use and potential consequences, CHAI has not published specific privacy, security, or human-review requirements for the five use cases. Data usage is another critical unknown; the programme may involve identifiable records, de-identified information, or synthetic data, but this has not been confirmed. Furthermore, the programme does not outline minimum staffing, infrastructure, or cybersecurity requirements for participating jurisdictions.
### **FAQ Section**
**Q: What is the PULSE programme?**
A: PULSE stands for Public Health Use Case and Learning Scaling Engine. It is a new programme designed to test generative AI tools within US public health departments. It is a collaboration between the Coalition for Health AI (CHAI), OpenAI, Anthropic, and Accenture.
**Q: Who is involved in the programme?**
A: The programme is led by the Coalition for Health AI. OpenAI and Anthropic are donating 10 enterprise licences for up to 2,000 practitioners. Accenture is overseeing onboarding and playbook development.
**Q: What will the AI be used for?**
A: The AI will be tested in five key areas: biosurveillance and drug-wave prediction, mapping social determinants of health, improving operations and efficiency, public communications and translation, and automating clinical data retrieval.
**Q: When will the trials begin?**
A: The pilot programmes are scheduled to begin in autumn 2026.
**Q: What about privacy and governance?**
A: Specific governance, evaluation metrics, and human-review protocols have not yet been defined. The programme will rely on frameworks like NIST’s AI Risk Management Framework, but CHAI has not published its own detailed requirements for these specific use cases.
### **Conclusion**
The PULSE programme represents a pioneering effort to harness the power of generative AI for the public good. By providing enterprise-level tools to a select group of public health agencies, the initiative aims to create scalable playbooks that can guide future adoption. While the potential for improved efficiency, better communication, and advanced data analysis is immense, the success of the programme will hinge on robust governance, clear evaluation standards, and unwavering attention to privacy and security. As the trials commence in 2026, the findings from the PULSE programme will be crucial in determining how AI shapes the future of public health in the United States.
