New study explores effectiveness of AI chatbots in HIV prevention
Novel research evaluates accuracy of AI-driven chatbots
- 3 min. read ▪ Published
New research conducted by the California HIV/AIDS Policy Research Centers (CHPRC) in conjunction with UC Berkeley School of Public Health explores the potential of artificial intelligence (AI) chatbots to aid in HIV prevention efforts.
The study, led by Marisa Fujimoto at UC Berkeley and the Northern California HIV/AIDS Policy Research Center, is entitled Evaluating AI Chatbots for HIV Prevention: An Assessment of Response Quality and User Tailoring and examines the ability of AI-driven chatbots to deliver accurate, engaging, and personalized health information to people from groups affected by HIV and community-based organizations.
As healthcare increasingly turns to digital solutions, this study provides critical insights into how AI can be leveraged to address HIV prevention in communities that may face barriers to traditional healthcare access. The research assesses not only the technical performance of these chatbots but also how well they cater to individual needs, offering an evaluation of both response quality and user-tailoring in a public health context.
Key Findings Include:
- High Response Accuracy, but Variable Clarity: AI chatbots can provide HIV prevention information and guidance that is accurate and neutral in tone across a wide range of HIV prevention topics, including pre-exposure prophylaxis (PrEP). However, some responses had a disjointed flow, lacked clear conclusions, and/or did not follow current best practices for use of non-stigmatizing HIV language.
- Personalized Engagement: Chatbots successfully simplified their responses when asked, but they largely did not tailor their responses to the needs of specific populations, such as transgender users or users in specific locations.
- Opportunities for Integration with Existing Public Health Services: When responses are reviewed and tailored by health professionals, AI chatbots may be a valuable tool for community-based organizations to enhance the efficiency and quality of service provision and to support the development of educational materials.
“New and innovative ways to enhance HIV care and prevention efforts are needed, especially to reach younger, tech-savvy groups who may turn to digital solutions for health information,” said Fujimoto. “Based on our results, we are cautiously optimistic about the use of AI chatbots for HIV prevention by individuals from communities affected by HIV, community organizations, and health providers. Chatbots are capable of providing reasonably accurate information with few access barriers and could be used best in conjunction with advice from health professionals to optimize information and provide referrals to services.
Nevertheless, our research also raises important questions about how to ensure that AI chatbots provide inclusive guidance that addresses the needs of communities disproportionately affected by HIV, like those seeking gender-affirming care.”
The research, funded by the California HIV/AIDS Research Program through the University of California Office of the President, was led by Fujimoto, Lauren Hunter, and Sandra McCoy from UC Berkeley School of Public Health, alongside Simon Outram and Laura Packel from UCSF.