A Proposal For Artificial Moral Pedagogical Agents

As artificial intelligence technologies continue to proliferate in all areas of contemporary life, researchers are looking for ways to make them safe for users. In the educational context, it is no different, from learning to explain-ing. This research presents an approach for value alignment in educational contexts, using artificial pedagogical moral agents (AMPA). We propose a top-down approach explaining why the bottom-up or the hybrid one, may would not be advisable when it comes to educational issues.