Effectiveness of artificial intelligence chatbots in providing sexual health information: an exploratory study

Generative artificial intelligence (AI) chatbots powered by large language models (LLMs) have demonstrated their potential in healthcare by offering relevant, user-specific information.1 Their versatility has been validated across domains like education and clinical decision support. In the domain of sexual health, accurate and stigma-free information is crucial for physical and psychological well-being. Misinformation or lack of access to credible information often contributes to anxiety, unhealthy behaviors, and delayed healthcare seeking. In many regions of world, cultural taboos further hinder open conversations about sexual health, making it difficult for individuals, especially youth to seek clarification from professionals.2 This emphasizes the need for anonymous, accessible sources of information. AI chatbots can fill this gap by offering personalized, stigma-free responses in a conversational style.3 This study aimed to assess the effectiveness of several such chatbots in addressing sexual health queries, evaluating their performance across key attributes such as accuracy, relevance, fluency, and coherence. The results of this analysis can inform their role in sexual health education, particularly in conservative settings.