True uncertainty and ethical AI: regulatory sandboxes as a policy tool for moral imagination

Abstract

We offer a complementary view to the ethical dilemmas discussed in the recent literature by pointing at the epistemological dilemma of true uncertainty, suggesting regulatory sandboxes as an apposite remedy. Using the exemplar case of the regulative sandbox for responsible artificial intelligence established by the Norwegian data protection authorities, we argue that regulative sandboxes have the potential of supporting the development of a more ethical AI through not only reducing uncertainty, but through nurturing moral imaginations.