For all that is written, claimed, criticised and published about content-based moderation, there is remarkably little which covers behaviour or conduct-based approaches. Despite this, many platforms large and small employ behavioural risk signals, especially to manage high-harm complex risks, and for the most proactive operations.

In this session, leading experts from within industry and academia will shed light on current approaches in 2025, efforts to increase transparency and usage, and debate what best practices look like; drawing on expertise applied to extremist groups, predators, and other high risk examples.