We propose that moral behaviour of artificial agents could (and should) be intrinsically grounded in their own sensory-motor experiences. Such an ability depends critically on seven types of competencies. First, intrinsic ...