Toward Human-Centered Responsible Artificial Intelligence: A Review of CHI Research and Industry Toolkits

As Artificial Intelligence (AI) continues to advance rapidly, it becomes increasingly important to consider AI's ethical and societal implications. In this paper, we present a bottom-up mapping of the current state of research in Human-Centered Responsible AI (HCR-AI) by thematically reviewing and analyzing 27 CHI research papers and 19 toolkits from industry and academia. Our results show that the current research in HCR-AI places a heavy emphasis on explainability, fairness, privacy, and security. We also found that there is space to research accountability in AI and build usable tools for non-experts to audit AI. While the CHI community has started to champion the well-being of individuals directly or indirectly impacted by AI, more research and toolkits are still required to address the long-term effects of AI on individuals, societies, and natural resources (human flourishing and sustainability).