Researching AI and Content Moderation: Trends, Tools and Methods


Eugenia Siapera, University College Dublin

Christian Katzenbach (HIIG/ZEMKI, University of Bremen) and Joao Carlos Magalhaes (HIIG)

Abejide Ade-Ibijola, University of Johannesburg


Giovanni De Gregorio


To register please click here.

We are pleased to announce a seminar series on AI, Hate Speech and Online Content Moderation with a focus on trends and challenges in the global south, with an emphasis on Africa. Covid19 has led to a rapid acceleration of automated content moderation and a reduction of human moderation. This has been problematic as algorithms have struggled to correctly interpret content leading to hate speech remaining online while other messages have been mistakenly identified as being ‘dangerous’. Furthermore, most algorithms have been developed and trained on content from the global north and have struggled to be sensitive to context, and even more so in countries in the global south where social media companies have been less engaged. These trends may have a particular impact in regions that are already affected by a multitude of destabilizing factors like climate change, political instability and economic uncertainty. This seminar series attempts to offer insight into some of the challenges faced by hate speech and automated content regulation on the continent.

This series is organized by Freie Universität Berlin, the Weizenbaum Institute, the School of Communications at the University of Johannesburg, and the Centre for Socio-Legal Studies’ Programme in Comparative Media Law and Policy (PCMLP) at the University of Oxford.  We are grateful for the support of the OxBer Research Fund and the ERC.