Title: Assessing and Communicating Risks of Generative AI in Public Health Information Ecosystems

Date: Monday, June 1, 2026

Time: 10 AM — 12 PM ET

Location: Virtual - Zoom link

Jiawei Zhou

Ph.D. Candidate in Human-Centered Computing

School of Interactive Computing

Georgia Institute of Technology

https://jiaweizhou.me/

Committee:

Dr. Munmun De Choudhury (Advisor) — School of Interactive Computing, Georgia Institute of Technology

Dr. Andrea G. Parker — School of Interactive Computing, Georgia Institute of Technology

Dr. Srijan Kumar — School of Computational Science and Engineering, Georgia Institute of Technology

Dr. Nicholas Diakopoulos — School of Communication, Northwestern University

Dr. Q. Vera Liao — Computer Science and Engineering, University of Michigan

Abstract:

Technology is increasingly shaping the way we interact with information, where both the quality of that information and the affordances of information technologies shape people's attitudes and decision-making. Generative Artificial Intelligence (AI) tools, such as large language models (LLMs), differ fundamentally from prior information and communication technologies that were primarily task-centric, by producing new content in a probabilistic manner and at scale. This generative nature is both the power and the pitfall of this new technology: it provides instant and scalable content, yet it can produce low-quality information (e.g., hallucinated outputs and oversimplified answers) as it is rapidly embedded into everyday systems, often without users' awareness of its role. This imbalance between adoption speed and public understanding raises pressing questions about how generative AI disrupts the ecosystems through which people seek, evaluate, and act on health information.

This dissertation argues that generative AI introduces ecological disruptions to public health information ecosystems that require contextually grounded risk assessment and literacy-centered risk communication to understand and address underlying harms. Situated within the context of a range of public health crises and challenges, my work employs computational, qualitative, and experimental methods, alongside deep-seated domain collaborations with experts in public health, communication, and natural language processing.

Across five studies amid the formative years of generative AI, this dissertation examines these disruptions at multiple levels. To assess risks, the thesis first showed that harm in low-quality information is jointly shaped by content composition and reader context, and that AI-generated content is structurally distinct from human-created content in ways that undermine existing detection and assessment solutions. The research then identified the risks of LLM adoption for public health information needs, spanning four ecological risk dimensions rooted in LLMs' generative nature. To examine how these risks are communicated, the work found that public discourse was not equipping people with a balanced view of AI capabilities and risks needed to navigate these technologies, and that explicit risk communication alone is insufficient to support informed AI use, with its effects contingent on the AI literacy users bring to it.

The dissertation reframes the problem of generative AI in public health information ecosystems in two directions: what AI does to the ecosystem by compressing distributed judgment into synthesized responses and simulating the signals users relied on to assess trustworthiness, and what the ecosystem fails to do in response by communicating risks to users who lack the conceptual foundation to act on them. Collectively, this research advances empirical grounding for more responsible adoption, communication, and governance of generative AI for health, with the aim of helping people safely and meaningfully engage with AI.