More than a tool: how generative artificial intelligence shapes employees' work goal progress through resource gains and losses

Purpose Generative artificial intelligence (AI) has been increasingly integrated into employees' daily workflows. This integration fosters sustained human-AI interaction, gradually shifting employees' primary interaction targets from coworkers to generative AI. However, prior research has primarily portrayed AI as a task-oriented “tool” and emphasized its instrumental role in supporting employee functioning, overlooking its emerging role as a relational “partner” in evolving interpersonal dynamics, and the implications this shift holds for employees' emotional experience and work goal progress. Design/methodology/approach Drawing on the cognitive-affective processing system (CAPS) framework and conservation of resources (COR) theory, we conducted a three-phase field study and an experimental study to theorize and test the double-edged effects of generative AI usage on employees' work goal progress. Findings Generative AI usage initiated a resource gain spiral by reducing emotional labor in the form of surface acting, thereby enhancing employees' work goal progress. Conversely, it also triggered a resource loss spiral by evoking organizational nostalgia (i.e. emotionally immersed in and mentally preoccupied with the past), which hindered employees' work goal progress. Furthermore, we found that attachment avoidance and attachment anxiety amplified the positive and negative impacts of generative AI usage, respectively. Originality/value Our study is among the first to examine the relational role of generative AI as a “partner” in workplace interactions and its double-edged impacts on employees' work goal progress. By integrating the CAPS with COR, we reveal that generative AI usage can both reduce emotional labor (resource gain) and evoke organizational nostalgia (resource loss), with these effects contingent on employees' attachment styles.