Abstract:Background: In Malaysia, HIV transmission among men who have sex with men (MSM) is rapidly increasing, fueled by persistent stigma, discrimination, and limited access to nonjudgmental healthcare. Leveraging artificial intelligence (AI) chatbots may offer innovative, scalable solutions to promote HIV prevention services in these hard-to-reach populations. Objective: This study aimed to evaluate the usability of a web-based AI chatbot embedded within the MYHIV365.com platform to promote HIV testing and related services among MSM in Malaysia. Methods: An observational study was conducted from August 2023 to March 2024 among 334 MSM, recruited through community outreach and social networking apps. The AI chatbot provided information on HIV testing, PrEP, mental health, and MSM-friendly clinics, and facilitated the ordering of HIV self-testing kits. Usability was assessed using predefined metrics, including interaction time, speech bubble count, topic selection, satisfaction score, and unanswered queries. Data were analyzed using descriptive statistics in R software. Findings: The chatbot recorded 393 interactions. Most users (91%) interacted once, with interaction times ranging from under 1 to 31 minutes. The chatbot facilitated 260 HIV self-test kit orders and provided clinic details in 93 interactions. Satisfaction was high (mean score = 1.3/1.1; 92.5% thumbs-up), and the mean recommendation score was 8.0/10. While the chatbot successfully answered 92% of queries, 8% remained unaddressed due to algorithm limitations. Conclusion: The AI chatbot demonstrated high usability and potential to complement HIV prevention efforts in Malaysia. To maximize its impact, future iterations must integrate more sophisticated AI models for personalized responses and build trust through stronger privacy protections. Linking the chatbot with health systems is essential for reducing stigma and increasing healthcare access for MSM.
Keywords: Artificial Intelligence and Healthcare, MSM, Artifical Intelligence, chat bot, health care, HIV
