Beyond the surveillance threat: when and how human-AI collaboration mitigates or magnifies employee privacy concerns

Abstract

Does human–AI collaboration necessarily heighten employees’ concerns for information privacy? Drawing on cognitive appraisal theory, this study challenges deterministic accounts of AI and privacy by proposing a dual-appraisal mechanism. Using a mixed-methods design (time-lagged survey and interviews), we show that human–AI collaboration simultaneously activates challenge and hindrance appraisals, producing divergent privacy responses. Challenge appraisals lead employees to interpret data practices as enabling performance and development, thereby reducing concern for information privacy. In contrast, hindrance appraisals frame AI as surveillance-oriented and autonomy-reducing, increasing privacy concerns. Importantly, AI intrusiveness weakens the challenge–privacy link while strengthening the hindrance–privacy pathway. Overall, the study develops a process-based model of workplace privacy, offering implications for responsible AI design and demonstrating that privacy concerns are cognitively constructed through appraisal dynamics rather than technological features.