Data mining creates a totalizing datification of reality. It blurs the borders between sensitive and non-sensitive data and between personal and non-personal data. It generates a classifying society in which associative and probabilistic analytics transform any runaway data into silent and unexpected determinants of decisions. In this society, the processing of a dataset does not relate any more to an individual but rather to the category/model in which the single data subject is classified.
Notwithstanding their ability to predict, data with deteriorating indirect connections to the targeted individual are often at the basis of these classifications. How compliant are they with the GDPR? In this context, should we search for explanations of how AI works, delve into the actual causality of things in real life or just be satisfied with probability? Can the GDPR help? The interdisciplinary discussion will focus on the following questions:
• What should be the role of causality, probability and explainability in the data driven society?
• Are data protection by design and by default useful in the classifying society?
• Is the GDPR the 21st century antidiscrimination law?
• Can data driven business models be reconciled with data protection?