Decision scoring without output rationales poses unacceptable risks for automated AI systems

Moral principles and responsible-use guidelines shaping how automated systems are built and deployed.