Dynamic, Community-Specific Moderation Models: A Comparative Study Across Federated and Mainstream Platforms

Building on Ermoshina & Musiani (2022)’s call for federated, socio-technical moderation models and Aitkenhead et al. (2024)’s successful co-creation in AutSPACEs, this research would systematically compare and prototype “living” moderation policies across a spectrum of platforms (centralized like TikTok/Facebook vs. federated like Mastodon or custom communities). It would involve ethnographic and design-based research with communities to iteratively craft moderation guidelines, enforcement mechanisms, and feedback loops that can be adapted over time—perhaps using modular policy frameworks or AI-assisted “policy suggestion engines.” Unlike most current systems, which are static and top-down, this approach centers on adaptability and user agency, and directly tests how federated governance can balance safety, autonomy, and inclusion. The result could be a toolkit or set of principles for any online community to implement context-sensitive, evolving moderation—potentially transforming the landscape for marginalized and niche communities.

References:

1. How to co-create content moderation policies: the case of the AutSPACEs project. Georgia Aitkenhead, Susanna Fantoni, James Scott, Sophia Batchelor, Helen Duncan, D. Llewellyn-Jones, Callum Mole, Otis Smith, Martin Stoffel, Robin Taylor, Kirstie Whitaker, Bastian Greshake Tzovaras (2024). Data & Policy.
2. Safer spaces by design? Federated architectures and alternative socio-technical models for content moderation. Ksenia Ermoshina, F. Musiani (2022).