TU Delft to Host Lecture on Cultural Bias in AI Models
Delft, Tuesday, 19 August 2025.
Joel Fischer will discuss the vital issue of cultural representation and bias in AI at TU Delft on 28 August 2025, focusing on ethical implications for technology use.
Addressing Cultural Bias in AI Systems
Joel Fischer, a Professor of Human-Computer Interaction at the University of Nottingham, will deliver a pivotal lecture at TU Delft on August 28, 2025. The lecture titled ‘Responsible AI and the Issue of Cultural Representation in Text-to-Image Model Outputs’ aims to reveal inherent biases in current AI systems that disproportionately favor representations of Western cultures in their outputs. This discourse is especially relevant as AI models, including Text-to-Image (T2I) systems, are increasingly utilized in diverse public-facing sectors such as education, media, and marketing [1][2].
The Role of Cultural Representation in AI
Fischer’s research underscores the over-representation of Western cultures in AI model outputs while minority cultures risk misrepresentation. To explore these cultural discrepancies, Fischer employs a combination of literature reviews, online workshops with participants of various cultural backgrounds, and evaluations of actual T2I model outputs. Through these methods, he intends to understand how these outputs align or misalign with cultural expectations across different demographic groups [1].
Implications for AI Stakeholders
The upcoming lecture will delve into broader implications for various stakeholders including developers, designers, end-users, and researchers. Fischer will link cultural sensitivity in AI to broader efforts of developing new assessment techniques that emphasize ethical considerations like responsibility, openness, and cultural sensitivity. This approach is aimed at promoting transparency within AI technologies and aligning them with ethical AI practices [1].
TU Delft’s Commitment to Responsible AI
Located in Delft, Netherlands, TU Delft is known for its commitment to addressing global challenges through education and innovation. This lecture, moderated by Himanshu Verma, an Assistant Professor at the Faculty of Industrial Design Engineering, represents part of TU Delft’s broader initiative to encourage discourse and development in responsible AI practices. TU Delft has been an advocate for ethical AI, and this event underscores its active role in facilitating discussions that could influence future advancements in AI technology [2][3].