Explainability, Human Oversight, and Procedural Justice in AI-Assisted Promotion Decisions: An Integrative Review for Chinese Organizations

AI tools are increasingly used to support internal talent decisions, yet promotion decisions pose a distinct governance problem because they involve future potential, not only past performance. Existing research has concentrated on recruitment screening or model accuracy, while the combined role of explainability, human oversight, and procedural justice in promotion contexts remains less settled. This paper develops an integrative review of research across human resource management, information systems, human-computer interaction, and AI governance to examine how managers and employees may respond to AI-assisted promotion decisions, with particular attention to Chinese organizations. Four conclusions emerge. First, explanations can improve perceived transparency, but they do not automatically protect users from poor AI advice. Second, human oversight only adds value when managers have both the authority and the criteria to question model output. Third, fairness in promotion decisions depends on voice, correctability, relevance of data, and accountability, rather than on statistical performance alone. Fourth, the Chinese regulatory context places additional emphasis on transparent and fair automated decision-making, which makes documentation, review, and appeal mechanisms especially important. On that basis, the paper proposes a practical framework for responsible AI-assisted promotion decisions built around data governance, interpretable evidence, structured human review, and employee contestability. The central argument is that organizations should not aim for uncritical trust in AI. They should aim for disciplined, reviewable, and job-relevant use.