A Framework for Integrating AI-Powered Systems to Mitigate Bias Risk in HRMFunctions

Integrating AI-powered systems in human resource management (HRM) functions offers advantages in enhancing decision-making, automating processes, and improving efficiency. However, these promises are associated with progressively hidden patterns and automated decision processes that may bake in existing biases. Despite the recent surge of research in this emerging topic, there remains a substantial gap between Artificial Intelligence's (AI) potential and practices in HRM, particularly in addressing diversity, equity, and inclusion (DEI) concerns within AI-driven HRM systems. This paper aims to contribute to filling this gap by reviewing and analysing existing literature in this domain and addressing two fundamental questions: what types of bias sources may be generated by AI-powered tools that could affect the DEI efforts, and second, what approaches and practices can reduce bias in AI-powered systems in HRM functions. The literature synthesis resulted in a conceptual framework outlining approaches to mitigate bias in AI-powered HRM systems across four key dimensions. Finally, the paper emphasises the need to overcome discrimination challenges posed by HRM innovations and optimise AI tools to enhance DEI efforts within the HRM function.