Insights into suggested Responsible AI (RAI) practices in real-world settings: a systematic literature review

We argue that AI-enabled systems have significant societal benefits, but only if they are developed, deployed, and used responsibly. But how can we ensure Responsible AI in real-world settings?

We systematically review 45 empirical studies in real-world settings to identify suggested Responsible AI practices to ensure that AI-enabled systems uphold stakeholders' legitimate interests and fundamental rights. Our review identifies 11 actionable Responsible AI practices and emphasizes the crucial role of stakeholders in high-risk AI deployments. We distill the findings into practical, how-to guidance for Responsible AI leads and practitioners, spotlighting best practices to date.