AI Chatbots vs. Traditional Sources: Dental Health Literacy and Confidence Among Dental Patients-A Cross-Sectional Study

Background and objectives: Given the increasing use of artificial intelligence (AI) tools for dental health information, this study compared patients' perceived health information literacy for AI chatbots, measured using the Artificial Intelligence-eHealth Literacy Scale (AI-eHEALS), with traditional health information literacy, measured using the Traditional Health Information Literacy Scale (THILS) and examined how literacy and confidence influence information source preferences. Materials and methods: A cross-sectional study was conducted at King Saud University Dental Hospital in Riyadh, Saudi Arabia, collecting responses from 474 adult dental patients using 8 validated items from AI-eHEALS and THILS. Domain-wise scores were calculated, and assumptions for parametric testing were verified. Group differences were analysed using independent t-tests. Structural equation modelling (SEM) examined pathways linking literacy domains, confidence, and information-source preferences, with perceived knowledge and evaluation items combined into a single AI-eHEALS literacy construct. Results: The THILS scores were significantly higher than those on the AI-eHEALS across all domains (p < .001). AI-eHEALS scores were positively associated with prior AI use, male gender, and employment status, while THILS showed few demographic links. SEM indicated that combined literacy (knowledge + evaluation) predicted confidence in using AI tools (β = 0.62). Usefulness items showed weak loadings and were excluded, improving model reliability. Participants reported moderate familiarity with AI chatbots; however, they demonstrated higher scores and greater agreement with traditional sources. Conclusions: The findings indicate that traditional dental health sources remain preferred over AI chatbots for health information, even with improved digital literacy. The study provides insight into patient perceptions of AI tools in dental care and suggests a need for enhanced AI literacy education and transparent communication about AI capabilities and limitations in health information contexts.