Abstract

The essay examines current discussions on how algorithms work from a normative view point. In particular, the attention is drawn to both the right to explanation enshrined in Art. 22 (3) of the European Data Protection Regulation (GDPR) and the principle of “explainabilityµ discussed in today’s institutional initiatives on the ethical principles of artificial intelligence (AI). The intent of the analysis is twofold. On the one hand, the aim is to show how, for any kind of knowledge of the algorithm or calculation program at stake, there is a sufficiently clear legal framework to establish liability in the event of damage. On the other hand, the goal is to specify which new principles and rights should complement the current regulatory framework, in order to address the ethical and legal challenges that algorithms pose in terms of knowability, explainability and responsibility for their design and use.