Universities have increasingly adopted AI-driven performance management systems (AI-PMS) to assess academics' contributions to teaching, research, and administration. However, these systems often rely heavily on quantifiable metrics, which may not fully capture academic freedom, mentorship quality, disciplinary differences, or context-specific professional performance. This reliance raises concerns about perceived unfairness, demotivation, strategic compliance, and resistance among academics. Drawing on organizational justice theory, a qualitative study, using stratified focus group interviews, was conducted to explore academics' perceptions of AI-PMS fairness, associated risks, resistance behaviors, and redesign needs within a public higher education setting in China . Eight focus groups, comprising six academics per group across different ranks and disciplines, participated voluntarily in the semi‑structured interviews. The focus was on specific AI-PMS features that need redesigning, perceived procedural and distributive injustices, emerging resistance behaviors, and recommendations for fairer system design. Reflexive thematic analysis was used to identify patterns of perceived unfairness and resistance. The analysis revealed education‑specific "fairness by design" principles, including algorithmic transparency, human‑in‑the‑loop review, appeal mechanisms, contextualized evaluation, and participatory design involving academic staff. These principles extended organizational justice theory to human‑AI interaction. The findings can assist public higher education institutions in designing more legitimate, transparent, and motivating AI‑PMS, thereby supporting Higher Education sustainability in line with SDG 4.

JEL Codes:

Keywords: AI-Driven Performance Management Systems; Organizational Justice; Academic Resistance; Fairness By Design; Higher Education Sustainability.