CEPE/IACAP Joint Conference 2021 - Abstract Submission and Registration

Session

Algorithmic Discrimination

6

Not scheduled

61. All variables are not created equal.

Individual Paper

The desire to systematise and quantify social science data for use in AI systems is growing. This article argues that social science data often cannot be converted into accurate categories on which algorithms work best. The literature raises this concern in a general way. Deeks (2018) notes that legal concepts, such as ‘proportionality’ cannot be easily converted into code. He notes that “The...

9. Informed Consent and Algorithmic Discrimination - Is giving away your data the new vulnerable?

Dr Hauke Behrendt (University of Stuttgart), Dr Wulf Loh (University of Tübingen)

Individual Paper

With the development and implementation of algorithm driven decision-making procedures based on aggregated data collected from individual users, several moral challenges arise. In this paper, we explore the connection between voluntarily sharing or selling your data on the one hand, and the dangers of automated decision-making based on big data and artificial intelligence on the other. We call...

68. Perception vs. reality of neutral AI

Individual Paper

Current examples show that the use of Artificial Intelligence (AI) can lead to discrimination against (mostly already marginalized) groups in society (Krauß, 2018), contradicting the demands for human-centered AI use. There are various sources of discrimination that can be found in every stage of development and application. AI-biases are difficult to predict ex ante, as well as difficult to...