Beyond functionality: how psychological needs shape affordance perception and motivated engagement in human-AI interaction

Users increasingly engage with AI systems that are technically capable yet psychologically dissonant, empowering in some moments, alienating in others. Existing Human-AI Interaction (HAI) research explains what AI systems can do but offers limited insight into why users engage or disengage at a motivational level. Drawing on Self-Determination Theory (SDT), Affordance Theory, and the Needs-Affordances-Features (NAF) framework, this paper develops an integrative conceptual model of motivated human-AI engagement. We propose that users’ psychological needs for autonomy, competence, and relatedness function as perceptual filters that shape which affordances they recognise and act upon in AI systems. These perceived affordances, moderated by user and task characteristics, determine the degree of psychological need satisfaction or frustration experienced during interaction, outcomes that in turn reinforce or undermine sustained engagement in a dynamic motivational loop. The framework advances HAI theory by recentring attention on human psychological experience rather than system functionality, and offers design implications for AI interfaces that support intrinsic motivation, user well-being, and meaningful human-AI collaboration.