Bournemouth University Higher Education Corporation and Clicka Servicing Company Limited KTP 24_25 R1

ClickASnap’s image-sharing platform is using artificial intelligence to automatically flag problematic user photos and help human moderators respond faster. The problem is scale. As a platform grows, human moderators cannot review every image uploaded. Offensive, illegal, or harmful content can slip through, or legitimate posts get wrongly removed. Current automated systems often struggle with nuance—misreading satire, art, or cultural context. This project aims to train AI to better understand still images and their captions, so it can accurately identify content that needs human review and even suggest appropriate responses. If the AI works as intended, moderation could become both faster and fairer. The platform would catch harmful content more reliably while reducing the number of innocent posts flagged by mistake. For users, this means a safer, less frustrating experience. For the company, it means fewer moderators stuck reviewing thousands of harmless photos. The approach could also be adapted by other image-sharing platforms facing the same moderation bottleneck—a quiet but essential part of keeping online spaces functional.