The Generative AI Dilemma in Knowledge-Intensive Teams

In knowledge-intensive teams, collaborative interactions among teammates have historically served as the primary means through which members access task-relevant information, advice, and feedback. Generative artificial intelligence (AI), however, now offers members an alternative source of these resources. We develop a multilevel theory that proposes team members who are proficient with generative AI are more likely to perceive the resources it can offer as functionally equivalent to those they can access through their teammates. In turn, we propose that some members will shift their knowledge-seeking interactions to generative AI. As this substitution behavior accumulates within a team, the team’s collaboration network is less likely to foster the critical emergent states that support team effectiveness. The key insight of our theory is a dilemma for knowledge-intensive teams: generative AI can boost team performance by helping members to work more efficiently, but a lack of collaboration among teammates can offset these gains and undermine team satisfaction.