Abstract: In January 2025, ahead of the second term of Trump’s presidency, Meta owner Mark Zuckerberg announced that the company would discontinue fact checking operations in the United States and introduce significant changes to its content moderation policies. The announcement has been met with severe criticism from fact-checking organizations, researchers, and civil society activists around the world, as it could set a dangerous precedent for corporations to backtrack on public responsibilities of oversight. Against these recent developments, this talk will examine the challenges of moderating problematic content on social media, and the promise and limitations of AI technologies in detecting contentious forms of speech. Drawing on the experiences of the AI4Dignity project, the talk will discuss 'ethical scaling' as a process model and a normative principle for AI-assisted content moderation and broader imaginings of AI. The talk will be based on the article, 'Ethical scaling: Extreme speech and the (in)significance of artificial intelligence' published in Big Data & Society, which won the International Communication Association Outstanding Article award in 2024.
Speaker: Sahana Udupa is Professor of Media Anthropology at the University of Munich (LMU Munich), Berkman Klein Fellow at Harvard University, and founder of the Center for Digital Dignity. She has published widely on political cultures of social media, AI and content moderation, decoloniality, news cultures, and platform governance. Udupa is a recipient of the Francqui Chair award in Belgium and European Research Council (ERC) grant awards. In 2024, she was named a Fellow of the International Communication Association. She is currently leading a new multinational, five-year project on small social media platforms. Udupa conducted her doctoral research at NIAS.