• News / 
• Events

Lecture Series of the Diversity & Gender Equality Working Group: AI, Discrimination & Stereotypes

On Thursday, May 21, 2026, the ComAI research group’s Diversity & Gender Equality Working Group launches its new internal Lecture Series with a talk by Katharina Mosene (Leibniz Institute for Media Research | Hans Bredow Institute).

Under the title AI, Discrimination & Stereotypes, Mosene examines how communicative AI systems – particularly Large Language Models (LLMs) – do not merely process information but actively reproduce and shape social meanings, narratives, and stereotypes. Drawing on an intersectional feminist perspective, she shows how existing power structures and forms of discrimination embedded in historically grown training data manifest in generated texts – for instance in gender-specific attributions, normative conceptions of work, or stereotypical representations of marginalized groups. Central to the talk is the question of whether LLMs can be understood as “communicative infrastructures” that not only reflect inequality but normalize it through everyday interactions. The talk concludes by discussing approaches to a discrimination-critical design and use of communicative AI.

Katharina Mosene is a political scientist at the Leibniz Institute for Media Research | Hans Bredow Institute (HBI), where she coordinates research and event partnerships and serves as the institute’s equality officer. Her scholarly interests focus on intersectional feminist approaches to internet policy, intelligent technologies, and internet governance. She is also co-project coordinator of the Human in the Loop research project at the Humboldt Institute for Internet and Society (HIIG) and a founding member of netzforma* e.V., an association for feminist internet policy.

The Lecture Series of the Diversity & Gender Equality Working Group aims to provide insights and impulses on questions of diversity and gender within ComAI research. We look forward to the talk and an engaging discussion!