Human-Autonomy Teams (HATs), which combine human members with autonomous systems such as artificial intelligence, encounter unique interaction difficulties when operating in dynamic environments. These hurdles often prevent effective adaptation and reduce overall potential. Yet, existing team theories do not fully capture HATs’ distinctive nature, and research remains fragmented across disciplines like human-machine interaction, offering limited guidance for both theory and practice. In response, this paper proposes enhancing current team frameworks by adopting a comprehensive perspective that accounts for the relationships among team members, tasks, and situational factors. Drawing on insights from multiple fields, it identifies four guiding principles—predictability, observability, plannability, and directability—as emergent states that anchor the cognitive and behavioral dimensions of managing adaptability in dynamic and interdependent contexts. By detailing the requirements for effective adaptation, the paper shows how HATs’ unique characteristics constrain their capacity to adjust. Finally, it explores the implications of these principles and presents concrete strategies for future research, training, and design.
