Protecting vulnerable and underserved AI users | IASEAI '26

How do we evaluate AI systems in real-world contexts—and who is most affected when they fail? This breakout session from IASEAI 2026 explores the challenges of auditing, evaluating, and governing AI systems across domains including mental health, advertising, generative media, and cultural representation. Across two segments, speakers examine bias, transparency, and trust in AI systems—highlighting both technical approaches and real-world consequences, particularly for vulnerable and underrepresented groups. Asmaa El Fraihi — An Experimental Framework for Auditing Ad Delivery Responsiveness to Psychological Traits Zeynep Tufekci — LLM Spirals of Delusion: A Benchmarking Audit Study of AI Chatbot Interfaces Nicholas Judd — Independent Clinical Evaluation of General-Purpose LLM Responses to Signals of Suicide Risk Sachin Pendse — The Agony of Opacity: Foundations for Reflective Interpretability in AI-Mediated Mental Health Support Huichan Seo — Exposing Blindspots: Cultural Bias Evaluation in Generative Image Models Smiljana Antonijevic Ubois — Addressing Historical and Contemporary Inequalities in Low-Resource Languages through the Lens of Serbian Yongxu Sun — VEAT Quantifies Implicit Associations in Text-to-Video Generator Sora and Reveals Challenges in Bias Mitigation Samuel Hill — Adaptive Trust Infrastructures for AI: From an AI ClinicalTrials.gov Framework to a Human-AI Workforce 📍 Recorded at UNESCO Headquarters, Paris #AI #AISafety #AIGovernance #AIpolicy #IASEAI #ResponsibleAI