Effective Moderation of Social Media to Curb Genocidal Content
The 2020-2022 Tigray war is reported to be the deadliest armed conflict of the 21𝑠𝑡 century, with an estimated 600 to 800,000 documented deaths and more than 100 thousand victims of rape as a weapon of war. Social media platforms were instrumental in spreading genocidal content during the conflict, and failure to effectively moderate hateful content resulted in the murder of civilians. This work investigates the expertise and processes required to effectively moderate such genocidal content, and compares these findings to the expertise and processes prioritized by social media platforms.
Read the paper and watch the 10-minute video below.
Read our paper for ACM CHI 2025
Social media platforms played a significant role in spreading genocidal content in the 2020-2022 Tigray war, where the deadliest genocide of the 21st century was committed. While linguistic expertise is clearly needed to adequately moderate such content, we ask: What additional expertise is needed? Why and to what extent do experts disagree on what constitutes harmful content, and what is the best way to resolve these disagreements? What do social media platforms do instead?
A new escalation of conflict between Ethiopia and Eritrea
Social media platforms are spreading violent warmongering content encouraging all-out war between Ethiopia and Eritrea, again. We call on the international community to help stop it.
Policy is urgently necessary to enable social media research
DAIR's Dylan Baker on the difficult task researchers face when trying to acquire and analyze data about social media platforms. Many of these platforms make accessing their data for research onerous at best -- a problem he says better policy can and must solve.
Image: Jamillah Knowles & Reset.Tech Australia / Better Images of AI / People with phones / CC-BY 4.0
Notes on scaling social media data collection
Dylan's advice for more easily downloading and processing social media data when corporate platforms do little to make this process accessible to researchers.
The human cost of our AI-driven future
DAIR's Adio-Adet Dinika writes for Noēma about the hidden, exploited workers fueling the data and training needs of artificial intelligence.
Image: Hanna Barakat & Archival Images of AI + AIxDESIGN / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/