Living health-promotion campaigns for communities in the United States: Decentralized content extraction and sharing through AI

Even though health-promotion campaigns can elicit behavioral change among constituents, these initiatives are generally implemented through expensive, centralized, unsystematic, and time-consuming efforts led by creatives and officials in federal and national agencies. Can advancements in AI provide systematic methods that generate living health campaigns out of social media posts generated by communities? Here, we report the success of an innovative method to automatically select actionable HIV prevention and testing messages from decentralized content on social media (e.g. X [formerly Twitter]). The method was assessed through computational methods, an online experiment with men who have sex with men, and a field experiment involving public health agencies and community-based organizations with jurisdiction in 42 counties in the United States. The computational analyses showed that the method is computationally successful. The results of the two experiments indicated that the resulting messages are perceived as more actionable, personally relevant, and effective, and the messages are six times as likely to be posted by agencies in United States counties.