Generative artificial intelligence (AI) is rapidly transforming knowledge-intensive work, yet its impact on team effectiveness remains unexplored. This article presents a multilevel theory articulating the process by which generative AI proficiency shapes member autonomy, task interdependence, and ultimately team effectiveness. Our theory proposes that members who are more proficient with generative AI experience a novel form of workplace control – information autonomy – which enables them to operate more independently of other team members. We theorize that members who experience information autonomy as a result of working with generative AI tools may shift some or all of their work-related dependencies from fellow team members to generative AI – a phenomenon we term dependence displacement. Although the productivity gains afforded by generative AI may boost team performance, we propose that generative AI may alter interaction patterns on the team in a way that negatively shapes team viability and satisfaction. Our theory contributes to the teams literature by examining how generative AI shapes interdependence within teams and to the job design literature by introducing information autonomy as a dimension of member-level control.
