Samenvatting

The idea of artificially intelligent systems committing crimes is by no means new. ‘Bad robots’ that rebel against humans have been the subject of science fiction for decades. Nowadays, AI has indeed penetrated our lives, and is here to stay. In this perspective, AI works as a stress test for classical criminal law constructs such as the concept of act and culpability. It is in this perspective that the question of ascribing criminal liability for ‘AIs going bad’ takes on relevance.