This paper aims to examine a practical risk in AI-supported human resource management: line managers may remain formally responsible for people decisions while relying too heavily on system-generated scores, rankings and recommendations. It argues that “human in the loop” is not enough unless managers are clearly accountable for how they use AI-supported information.
This practice-oriented viewpoint draws on research on AI in HRM, algorithmic management, workplace fairness and human judgement, and translates these debates into a practical playbook for senior HR leaders.
AI-supported HR tools can improve speed, consistency and evidence-informed decision-making. However, they can also blur responsibility when managers treat outputs as conclusions rather than prompts for judgement. The study proposes a five-step playbook: assign decision ownership, ask better questions, add context, account for overrides and answer employees in human terms.
HR leaders should govern not only AI systems, but also the managerial behaviours around them. This requires clearer decision rights, manager training, contextual review, documentation rules and employee-facing explanations.
The study reframes responsible AI-supported HRM as a practical management accountability issue. It offers senior HR leaders a simple framework for ensuring that AI strengthens, rather than replaces, human judgement.
