AI for Preparedness at University of Arizona Public Health and AI Summer School

In an era where false information can spread faster than the facts, getting the right message to the right person at the right time has become a big challenge in public health. At the Public Health & AI Summer School (June 2025 agenda) hosted by the University of Arizona, I had the opportunity to present how Wehealth is tackling this challenge using a human-centered, AI-powered approach. [Slides] [Event coverage].

The Problem: An Overwhelming Information Landscape

Our public health system is decentralized, with over 3,000 counties and territories each managing their own messaging and response. This decentralization can lead to:

• Duplication of efforts
• Inconsistent messaging
• Limited visibility for trusted sources
• Barriers to collaboration and coordination
• Underutilized but high-quality public health content

Worse, the information ecosystem itself has become more chaotic. According to the Lloyd’s Register Foundation World Risk Poll, false information is now the #1 concern among people globally. Yet, trusted public health voices still struggle to be heard.

At the same time, disparities in English language proficiency, access to digital tools, and healthcare resources make it harder for marginalized populations to receive and act on critical information—especially during emergencies like extreme heat events or pandemics.

A New Approach: The AI-Assisted Public Health Communication Loop

Wehealth is building what we call the AI-assisted communication loop—a 4-step system that strengthens trust, reach, and relevance in public health messaging:

1. Aggregate Verified Information

We ingest and normalize trusted content from sources like the CDC, NWS, and local health departments, using both structured APIs and unstructured web content.

2. Enrich It Using AI

We personalize the message using a combination of environmental data, local health indicators, and individual-level risk factors (e.g., age, chronic conditions, ability to cool down in heat).

This enrichment step includes:

• Translation into multiple languages
• Conversion to video and audio formats
• Customization for accessibility and cultural relevance

3. Amplify via Trusted Channels

We broadcast the enriched information via a wide range of communication tools:

• Mobile apps (iOS, Android, Web)
• SMS/Text messaging
• Email, web embeds, and 3rd-party systems

Our model, phAI (Φ), is a custom-built AI engine designed specifically for real-time, critical public health messaging with safety, personalization, and trust built in.

4. Listen and Guide

The loop doesn’t stop at broadcasting. We collect feedback, measure engagement, and help users take informed actions. For example, during a heat wave, a resident might receive a message that says:

“Stay indoors between 12–4pm. Check your medications for heat sensitivity. The nearest cooling center is at 123 Main Street.”

Guardrails, Safety, and Human Oversight

Using AI for public communication brings real risks—but also real opportunities. That’s why humans are always in the loop at Wehealth. Here’s how we ensure safety and trust:

• Expert Moderation: Every message, translation, and video generated by our AI is reviewed by trained public health professionals.
• Quality Assurance Automation: We run 24/7 monitoring and simulations to ensure messages are delivered properly, and any system issues are caught early.
• Audit Trails: All content changes are logged and reviewable, ensuring transparency and accountability.

Wehealth’s AI even checks its own outputs—flagging discrepancies, triggering alerts, and enabling continuous improvement.

The Tools: A Unified Platform for Public Health

Our solution isn’t just for the public—it’s also built for the people who serve the public.

• Wehealth Portal: A powerful dashboard for health departments to send alerts, manage content, run surveys, and view analytics. Available on web and mobile.
• Data Engine: A robust backend that connects to existing systems, ingests diverse data formats, and enables seamless integration across agencies.
• APIs and Integrations: Designed for extensibility, partners can build on our platform, query our AI engine, and integrate with their own tools.

Our goal is to make it easy, fast, and safe for public health agencies to communicate during high-stakes moments—while giving them the flexibility to customize for local needs.

A Call to Collaboration

Wehealth is more than a technology company, we’re building a public good in collaboration with forward-thinking public health departments, researchers, and civic leaders.

We invite you to join our expert community to participate in our monthly meetings, share feedback, and help shape the future of trusted public health communication.