An Ethical Framework for Guiding the Development of Affectively-Aware Artificial Intelligence

The recent rapid advancements in artificial intelligence research and\ndeployment have sparked more discussion about the potential ramifications of\nsocially- and emotionally-intelligent AI. The question is not if research can\nproduce such affectively-aware AI, but when it will. What will it mean for\nsociety when machines -- and the corporations and governments they serve -- can\n"read" people's minds and emotions? What should developers and operators of\nsuch AI do, and what should they not do? The goal of this article is to\npre-empt some of the potential implications of these developments, and propose\na set of guidelines for evaluating the (moral and) ethical consequences of\naffectively-aware AI, in order to guide researchers, industry professionals,\nand policy-makers. We propose a multi-stakeholder analysis framework that\nseparates the ethical responsibilities of AI Developers vis-\\`a-vis the\nentities that deploy such AI -- which we term Operators. Our analysis produces\ntwo pillars that clarify the responsibilities of each of these stakeholders:\nProvable Beneficence, which rests on proving the effectiveness of the AI, and\nResponsible Stewardship, which governs responsible collection, use, and storage\nof data and the decisions made from such data. We end with recommendations for\nresearchers, developers, operators, as well as regulators and law-makers.\n