<![CDATA[Community health literacy remains uneven because social media users frequently encounter complex, misleading, or poorly adapted health information. This study aimed to evaluate whether human-reviewed AI-generated social media campaigns improve health literacy more effectively than conventional campaigns and standard public health information. A multi-site cluster-randomized mixed-methods design involved 660 adults from twelve urban, suburban, and rural communities. Participants received an eight-week AI-supported campaign, a conventionally developed campaign, or standard information, with assessments conducted at baseline, post-intervention, and six-month follow-up. Results showed that the AI-supported campaign produced the greatest improvements in health knowledge, message comprehension, misinformation recognition, source evaluation, and preventive action. Health-literacy scores increased from 52.84 to 68.91 and remained at 66.72 after six months. Message comprehension, perceived relevance, and trust explained more of the intervention effect than exposure or engagement alone. The study concludes that AI can strengthen community health literacy when content is medically verified, culturally adapted, transparent, and supported by community participation. Campaign success should be evaluated through educational and behavioral outcomes rather than platform popularity alone. Human oversight remains essential for accuracy, equity, privacy, and public trust, while locally responsive design can improve access among populations with lower baseline literacy and diverse levels of digital competence. ]]>
Copyrights © 2026
