Can AI Assistants Help Address Public Health Questions?

A new study by researchers from the Qualcomm Institute at the University of California San Diego highlights the potential of AI assistants in providing accurate and actionable responses to public health questions. The study evaluated ChatGPT’s ability to address common health queries and found that it provided evidence-based responses to 91% of the questions. However, only 22% of the responses included referrals to specific resources, indicating a need to improve connecting users with appropriate help. The study team asked ChatGPT 23 public health questions belonging to four categories, namely addiction, interpersonal violence, mental health, and physical health. While ChatGPT’s responses mirrored the type of support that might be given by subject matter experts, they rarely made referrals to specific resources to help the questioner. According to the researchers, connecting people to trained professionals should be a key requirement of these AI systems and could substantially improve public health outcomes.