Workplace Relationships and Artificial Intelligence

Generative AI is now embedded in how coworkers, leaders, and organizations communicate, yet its effects on the relationships among them are still poorly understood. This review synthesizes recent research on how an individual’s AI use changes the way others perceive, judge, and behave toward them, and shows that these effects can be explained by established theories of human-human interaction, including attribution theory, computers-as-social-actors, the integrative model of organizational trust, social exchange theory, and resource-based theories. The evidence suggests that observers often read AI use as a sign of disposition rather than circumstance, judging AI users as less competent, motivated, warm, and authentic, and that disclosure does not repair the resulting loss of trust. These perceptions carry over into behavior: coworkers reciprocate less, share less knowledge, and show less prosocial behavior toward AI users, while AI-related job insecurity, loneliness, and work alienation further strain ties among colleagues. AI use also loosens moral standards and shifts how blame is assigned, making people more critical of others’ work but more lenient toward misconduct. These effects are conditional rather than uniformly negative. AI can reduce loneliness, improve coordination, and prompt employees to invest more in emotionally supportive ties, and relational buffers such as leader-member exchange, shared leadership, and supervisor support can offset its costs. The review concludes that AI’s costs to workplace relationships run mainly through indirect pathways, not through the technology itself. It outlines organizational implications, including clear norms for when and why AI is used and protection of socially valued tasks, and identifies priorities for future research: longitudinal designs, team- and network-level studies, intervention research, and more diverse samples.