Exploring the Future of Health Information Literacy through Generative AI Integration: A Pilot Study

This pilot study explores the integration of generative artificial intelligence (GAI) to expand health information literacy and accessibility within a linguistically and culturally diverse region. The project focuses on developing and deploying a ChatGPT-powered conversational agent hosted in the university’s health sciences library. The core aim is to create an interactive, user-friendly chatbot trained on high-quality, open-access health resources—such as MedlinePlus and other reputable databases—so that students, staff, faculty and other library users can obtain instant, tailored and reliable health information at point of need.

The initiative is structured across distinct phases. The initial development and training phase concentrates on curating a focused knowledge base and fine-tuning a conversational model to provide accurate, understandable health information while minimizing hallucinations and ensuring the use of authoritative sources. Training data selection emphasizes MedlinePlus and comparable open-access health databases to concentrate the chatbot’s responses on evidence-based content. Subsequent technical implementation includes integrating the chatbot with library infrastructure and designing an access point equipped with an interactive tablet to facilitate in-person and on-site engagement. This placement is intended to lower technological barriers for users who may not have ready access to personal devices or who benefit from guided interactions within a familiar library setting.

Equity and digital health literacy are central to the project’s objectives. The team intends to address language barriers and culturally responsive communication by tailoring the chatbot’s interface and response style for diverse user groups encountered within the region. By situating the tool within the health sciences library, the project couples technological access with opportunities for supplementary instruction and support from library staff and health information professionals. This dual approach aims to not only deliver information but also to scaffold users’ ability to find, interpret, and apply health information—core components of health information literacy.

Evaluation and iterative improvement are built into the pilot’s design. Planned assessment activities include monitoring usage metrics, collecting user feedback, and evaluating the accuracy, clarity and usefulness of chatbot responses. Insights from these evaluations will inform refinements to the chatbot’s training set, conversational design, and user interface. The pilot’s period of performance runs from June 1, 2024, through April 30, 2025, enabling a sustained implementation and an opportunity to measure short-term outcomes related to access and user experience.

Funded through NNLM’s Region 5 award mechanisms, this outreach-focused project situates a novel GAI application within a health sciences library context, aiming to model how conversational AI can support equitable access to authoritative health information. The pilot is positioned to generate practical lessons about integrating GAI into library services, enhancing digital health literacy, and addressing language and cultural diversity in health information delivery. Findings and practices from this work could inform broader NNLM initiatives and library-based strategies to responsibly deploy generative AI tools for public health education and information access.