ITeG lecture series: "Bias in data and algorithms "at work": old problems, new approaches."

In the ITeG lecture series "Digital Society - A Design Task", Prof. Dr. Bettina Berendt will give insights on the topic "Bias in Data and Algorithms "at work": old problems, new approaches" on February 8, 2023.

Bettina Berendt has been Professor of Internet and Society at the Technical University of Berlin since 2019, Director of the Weizenbaum Institute and Visiting Professor at the KU Leuven, Belgium. Her current research includes data science and critical data science, especially regarding privacy/data protection, discrimination and fairness, as well as AI and ethics, with a focus on textual and web-related data.

Summary of her presentation:
In current discussions, one often hears that Artificial Intelligence could have "bias", that the use of AI could lead to "discrimination". Laws and draft laws such as the GDPR and the draft AI regulation also mention these risks and aim to contain them. But what do we actually mean when we talk about "algorithmic bias" (etc.) and what can computer scientists and other stakeholders do about it? In this talk, I will use observed and (somewhat) fictional examples and an interdisciplinary perspective to illustrate how AI and other technology can systematically disadvantage people. I present structural patterns of discrimination and outline principles of countermeasures. It turns out that many of the seemingly "new", "technology-driven" problems are already familiar "old problems" and that human and machine decisions are intertwined. Finally, I give an outlook on how such an understanding can also contribute to using technology to reduce discrimination and disadvantage.

The lectures will take place online (via Zoom). You can find the link for the Zoom meeting and the meeting ID on the following website: www.uni-kassel.de/go/iteg-lectures