Presentation (CTI)

OBJECTIVES: Large language model (LLM)-based Generative artificial intelligence (GenAI) is transforming health communication, but might produce probabilistic outputs that are difficult to interpret and apply appropriately in healthcare contexts. Existing health literacy HL and AI literacy frameworks do not fully capture the competencies required for safe patient engagement with GenAI. Two key gaps are identified: a generative navigation gap (the ability to formulate, interpret, and iteratively refine AI interactions) and a clinical contextualization gap (the ability to appropriately translate AI outputs into condition-specific health decisions). This study aimed to conceptualise GenAI Health Literacy (GenAIHL) as a competency framework for safe, effective, and equitable patient engagement with GenAI.
METHODS: A narrative review and integrative framework synthesis were conducted, drawing on literature in health literacy, digital health literacy, and AI literacy to develop a conceptual competency framework.
RESULTS: GenAI health literacy (GenAIHL) is conceptualised as a multidimensional construct integrating established health literacy domains with key dimensions of GenAI literacy. The framework comprises four interrelated domains: 1) functional (technical understanding of how GenAI systems generate outputs and the importance of input quality), 2) interactive (ability to formulate clear, context-specific prompts and iteratively refine queries), 3) critical (capacity to appraise the reliability, consistency, and potential bias of AI-generated information), and 4) empowered (responsible and ethical use, including privacy awareness and appropriate reliance on professional care). Together, these domains reflect the capabilities required for patients to engage with GenAI as active participants rather than passive recipients. By addressing capabilities that influence patients’ ability to access, interpret, and apply AI-generated information, the framework provides a foundation for reducing inequities in GenAI use.
CONCLUSIONS: GenAIHL defines a patient-level capability layer necessary for safe, effective, and equitable use of LLM-based health systems. It extends existing literacy frameworks by incorporating generative interaction and clinical contextualization competencies.