Christiane Attig and Christiane Wiebel, "Beyond Tools: The Perception of AI as a Social Teammate in Human-AI Collaboration", 13. Fach­grup­pen­ta­gung der Fach­grup­pe Ar­beits-, Or­ga­ni­sa­ti­ons- und Wirt­schafts­psy­cho­lo­gie ge­mein­sam mit der Fach­grup­pe In­ge­nieur­psy­cho­lo­gie (AOWI 2025), 2025.

Human-AI interaction is often framed in terms of task efficiency, but users do not necessarily interact with AI systems as tools. Instead, they may perceive them as social partners with specified roles. The subjective ascription of roles in human-AI teaming can be understood as a multi-factorial, dynamic process, depending on task characteristics, AI functionality, and the user’s perception of the AI and the joint task. Consequently, to predict the users’ related expectations towards an AI system within a shared task, it is essential to consider the interplay between contextual factors (e.g.: properties of the AI system), and the users’ subjective perception, which is shaped by social cognition and motivational needs. Drawing from Self-Determination Theory, we propose that users evaluate AI agents based on their ability to support competence (effectively support task completion), autonomy (provide control and flexibility), and relatedness (engage in social behavior or fostering human-human connection). Based on integrated empirical findings, we will argue that an AI system which is perceived as fulfilling these needs is more likely to be experienced as a team partner rather than a mere tool, paving the way for joyful human-AI teaming endeavors.