This project aims to amplify the perspectives of underrepresented students, particularly those from marginalised backgrounds, by using AI to detect and analyse personal narratives of school bullying shared on social media platforms.

Using multilingual and multimodal foundation models, the project will identify patterns in the forms and drivers of bullying, including potential institutional factors affecting these cohorts.

Findings will be used to generate actionable, equity-centred insights to inform anti-bullying policy and strengthen student support systems.