ART-AI Seminar
We are pleased to have Alexia Gaudeul, a behavioural researcher at the EU Policy Lab and Ottla Arrigoni, a service designer for the public and third sector who are both at the European Commission join us for this ART-AI seminar entitled ‘Understanding the Impact of Human Oversight on Discriminatory Outcomes in AI-Supported Decision Making’.
This seminar will take place online on Zoom on Tuesday 16th July 2024, 12.15pm-1.15pm (BST). Elsa Zhong will be chairing. For joining instructions, please e-mail [email protected].
Title
Understanding the Impact of Human Oversight on Discriminatory Outcomes in AI-Supported Decision Making
Abstract
We present a large-scale study to assess the impact of human oversight on countering discrimination in AI-aided decision-making for sensitive tasks. We use a mixed method approach, including a quantitative experiment with HR and banking professionals in Italy and Germany (N=1411), and qualitative analyses through interviews and workshops with participants and with fair AI experts. The results show that human overseers were equally likely to follow advice from a fair AI as from a generic AI, and that human oversight did not prevent increased discrimination from a generic AI. Fair AI reduced gender bias but not nationality bias. Participants’ choices were neither more nor less responsive to their preferences when using an AI compared to when left on their own. Interviews with participants showed they prioritized their company’s interests over fairness and highlighted the need for guidance on overriding AI recommendations. Fair AI experts emphasized the need for a comprehensive approach when designing oversight systems.
Bio
Alexia Gaudeul is a behavioural researcher at the EU Policy Lab of the European Commission. She holds a PhD in economics from the University of Toulouse and worked at universities in the UK and Germany. Her research focuses on the economics of the Internet and of our digital lives. She currently investigates how biases affect our perception of advice from AI. She also works on digital product labelling and fighting misinformation.
Ottla Arrigoni is a service designer for the public and third sector, currently working on Design for Policy at the EU Policy Lab, at the European Commission. Ottla’s experience in China, UK and now Belgium, spans a series of topics, including AI, sustainability, housing, and healthcare. She believes in bringing participatory design, and system thinking in every project, developing interventions in context, and visualising complexity.
