About Forough.
Forough Poursabzi-Sangdeh is a Principal Applied Scientist in Microsoft's Office of Chief Scientific Officer. Her research sits at the intersection of AI, people, and society, drawing on computer science, psychology, social sciences, and human-computer interaction to identify and mitigate risks in AI systems. Recently, her work has centered on the psychological risks posed by AI and translating research findings into evidence-based policy and design guidance. Forough co-leads a cross-company initiative on the psychological influences of AI, with the goal of shaping the responsible development of AI chatbots by focusing on safeguarding user wellbeing and fostering healthy human-AI interaction.
Before beginning her work on the psychological influences of AI, Forough led responsible AI red teaming for several prominent AI models, including DALL-E, GPT-4, and Codex. She translated insights from this work into Microsoft's practices and guidance for red teaming generative AI systems, helping expand the discipline beyond traditional security testing to examine a broader range of responsible AI risks. You can read more about this work in Microsoft's blog post on its AI Red Team.
Forough holds a PhD in Computer Science from the University of Colorado Boulder, where she specialized in human-centered machine learning, and a BE in Computer Engineering from the University of Tehran.
Beyond research, Forough has been interested and active with efforts on diversity and inclusion in computing. In 2019, she was a co-organizer of the Women in Machine Learning (WiML) workshop at NeurIPS.
If you are curious, Forough is pronounced roughly as foe-roogh. The 'gh' is a soft, guttral sound made in the back of the throat, a bit similar to how 'r' is pronounced in French. It is a Persian name and it means brightness and light ✨
