This will be a presentation on how we build effective partnerships with local civil society organisations to address harmful online content, particularly in high-risk environments and on issues that require nuanced understanding of the local context. In a world of AI and machine learning, the presentation will focus on how we continue to incorporate the 'human' aspect into policy development and content moderation through civil society partnerships, and how we ensure that we continue to bring local context to our global Trust & Safety policies and processes. Meta's Trusted Partner Program brings together over 400 civil society organisations across 113 countries to address harmful content on Meta's platforms and better understand the context of the diverse communities in which we operate around the globe.