Advances in Artificial Intelligence (AI) reshape many aspects of human life creating new opportunities for value creation but at the same time posing significant threats. This paper posits that as AI systems increasingly influence societies, their development and use need to be responsible and aligned with human rights. Multiple principles and guidelines for responsible AI have been proposed during the last few years. These principles are frequently presented as lists with little coherence and no theoretical basis. This paper draws from Rawls´ “justice as fairness” theory to propose a hierarchical framework of ten interrelated principles for responsible AI. This framework can be used as a roadmap for organizations towards increasing their maturity for developing, deploying and managing AI responsibly.