Skip to main content
NORC at the University Chicago

Health Communication AI

Young woman reading message on her smartphone while standing in the living room. Her hair is dyed teal and her left arm has no forearm.
Combining the trust-building ability of digital opinion leaders with the scalability of AI to revolutionize health communication
  • Dates
    December 2023 – Present

Problem

Artificial intelligence is drastically changing the way we engage with health information.

The rise of a fragmented ecosystem, inaccurate information, a widening digital divide, and the emergence of artificial intelligence (AI) have introduced new complexities to health communication.

Public health is particularly affected as individuals increasingly turn to the internet and social media to make informed health decisions.

Increasingly, people are engaging with AI agents for information about health. Currently, about 17 percent of adults use AI at least once a month to access health information. That percentage increases substantially among youth, where it is estimated that over a quarter of U.S. teens use generative AI to seek such information.

Solution

NORC is spearheading Health Communication AI, a multi-pronged initiative that is examining the intersection of AI and health in today’s AI-driven communications environment—across multiple dimensions.  

This initiative is developing a portfolio of research and building the evidence base related to AI and health communication.  

First, we have developed an AI agent that we are evaluating for use in many different health contexts. By leveraging the extensive knowledge base of digital opinion leadership and integrating it with generative AI technology, our AI agent delivers relevant and empathetic health information when and where it is needed. With its capacity for social listening, sentiment analysis, and accurate generative response, our agent is poised to address critical gaps in digital health communication.  

We are also helping build an entirely new scientific agenda that includes the development of:  

  • New and updated theoretical frameworks  
  • Pilot and feasibility studies to measure the effectiveness of such an integrated AI system on accuracy, empathy, and user response  
  • Longitudinal studies to examine impact on health behaviors  
  • Implications and recommendations for policy, ethics, and privacy, as well as practical application 

Result

To address modern health communication challenges at scale, NORC is designing and leading several innovative projects under its Health Communication AI initiative.

This includes:

  • Collaborating with Amazon Web Services to develop a supervised model of online covid vaccine information.
  • Developing and testing an AI agent trained on evidence-based health information, which we are testing for a variety of topics including HPV vaccination and cardiovascular disease. 
  • Beginning a new project with Amazon Web Services to evolve our AI agent. 
  • Partnering with OpenAI to evaluate AI agents’ performance and develop benchmarks for key metrics such as accuracy, harm potential, and empathy. 
  • Collaborating with colleagues to examine cross-cultural use of AI and adolescent girls’ body image perceptions. 
  • Beginning to explore the impacts of AI engagement on youth mental health and develop supportive communication resources. 

As part of this work, we have shared our insights through journal articles and at conferences. Our hope is that the body of work that results from our Health Communication AI initiative fosters innovation in public health and medical communication and enables additional research to create an evidence base for AI use in health communication.   

Amelia Burke-Garcia, PhD, director of NORC’s Center for Health Communication Science, oversees the Health Communication AI initiative. She is also co-principal investigator, along with Rebecca Soskin Hicks, MD, fellow at NORC, on the AI Agent research.  

If you want to work with us as part of our Health Communication AI initiative, contact  healthcommai@norc.org. 

Project Leads

Resources

FAQ

Health Communication AI is an innovative approach that NORC is spearheading which aims to improve public health communication and combat inaccurate health information. It builds upon evidence from the opinion leadership knowledgebase, further extending recent work that has examined opinion leadership in the digital age by integrating emerging generative AI technology. As such, Health Communication AI creates personalized and empathetic interactions in order to enhance and scale the dissemination of evidence-based health information.

The project was inspired by the challenges faced in health communication, including increasing distrust of our public health and health care systems, the rapid and vast spread of inaccurate health information, and the difficulty of engaging with diverse communities at scale. Combining Amelia Burke Garcia’s research on social media influencers as the newest opinion leaders and Rebecca Soskin Hicks’ experience in AI advancements and digital products for health communication provides a unique opportunity to innovatively address these challenges.

Health Communication AI necessitates a robust research agenda—one that NORC is leading through its Health Communication Science work. This agenda includes conducting feasibility and pilot studies to better understand what works and what does not in this new environment, as well as longitudinal studies to examine long-term impacts on health decision-making and behaviors. It also includes work to unveil implications and recommendations related to policy, ethics, and privacy, as well as practical applications.

By leveraging AI's ability to analyze vast amounts of data and generate responses that are both accurate and tailored to an individual's concerns, Health Communication AI can provide credible, evidence-based information more effectively than traditional methods. This proactive approach helps counteract inaccurate health information by ensuring that accurate information is present in the conversation—when and where it is happening on social media and in digital spaces—and by making the information more accessible and engaging through empathetic and unbiased communication.

Health Communication AI is an integrated AI system, functioning as an AI agent that executes social listening, sentiment analysis, and provides text responses through a fine-tuned chatbot. The chatbot engages with individual users on social media and optionally through websites, the GPT store or apps.

Yes, the AI models can be trained on specific health topics, cultural considerations, and languages, making it adaptable to various public health contexts. The evidence base on which it is fine-tuned informs the chatbot’s development, ensuring it aligns with health communication best practices and up-to-date health and medical data. This flexibility ensures that the communication is relevant and effective across different populations and health challenges.

Ethical considerations include ensuring data privacy, avoiding bias in AI-generated content, and maintaining transparency about the use of AI in communications. Developing and implementing guidelines that safeguard individuals' rights and ensure the responsible use of AI technology are central to NORC’s work on Health Communication AI.

We are actively seeking partners who want to invest in the development of this idea, co-design and participate in pilot programs, provide expertise and feedback, and explore collaboration opportunities for deploying Health Communication AI within their networks or communities. Contact burkegarcia-amelia@norc.org for more information.