GRACE NSEABASI EKANEM

Background: Artificial intelligence (AI) chatbots are increasingly being used as source of health information, particularly among undergraduate students who are highly engaged with digital technologies. These tools provide instant, interactive, and personalized responses to healthrelated queries, which may influence health-seeking behaviors. One growing concern is their role in facilitating self-medication, defined as the use of medicines without consultation with qualified healthcare professionals. While AI chatbots may improve access to health information and empower individuals to make decisions, their unregulated use raises concerns about misinformation, inappropriate drug use, delayed diagnosis, and adverse health outcomes. Despite the increasing global use of AI technologies, there is limited evidence on how undergraduate students in Nigeria utilize AI chatbots in relation to self-medication practices. Understanding students’ knowledge, attitudes, and patterns of use is essential for informing public health interventions and policies. Methods: An analytical cross-sectional study was conducted among undergraduate students in Benin City, Nigeria. Data were collected using a structured, self-administered questionnaire adapted from UTAUT and related acceptance models that assessed socio-demographic characteristics, knowledge of AI chatbots, attitudes toward their use in health decision-making, and prevalence of their use in facilitating self-medication. Knowledge and attitude scores were computed and categorized into levels. Data analysis was performed using appropriate statistical software. Descriptive statistics such as frequencies and proportions were used to summarize variables, while inferential statistics, including chi-square tests, were used to examine associations between variables. Statistical significance was set at p < 0.05.Results: The mean age of respondents was 21.50±3.138 years, with females constituting the majority (78.4%). Awareness of AI chatbots was universal, and about four-fifth of respondents demonstrated good knowledge, with Gemini being the most correctly identified tool. Despite this high awareness, only a small proportion had received formal training on AI or chatbots. About seven-tenth of respondents expressed a positive attitude toward AI chatbot use, perceiving these tools as convenient and useful for obtaining quick health information, although concerns regarding reliability and safety remained common. The prevalence of AI chatbot use for self- medication was considerable, with nearly one-third of respondents reporting use for advice on symptoms, possible diagnoses, and treatment options. ChatGPT was the most commonly used chatbot for self-medication, followed by Gemini. Despite the prevalence of use, the frequency of chatbot utilization for self-medication was mostly occasional or rare. Sex and guardians occupation were significant predictors of good knowledge. Attitude toward AI chatbot use was a strong predictor of prevalence. Respondents with a positive attitude were significantly less likely to use AI chatbots for self-medication compared with those with a negative attitude (OR = 0.178, p < 0.001) Conclusion: Despite high awareness and good knowledge of AI chatbots among respondents, concerns about reliability and safety in self-medication persisted. About one-third had used AI chatbots, mainly ChatGPT, for self-medication. Knowledge, attitude, guardians’ occupation, and social media use significantly influenced utilization, highlighting the need for targeted health education, improved digital health literacy, and regulatory frameworks to ensure safe and responsible use of AI chatbots in healthcare decision-making.