AI-Mediated Exchange Theory

AI acts as more than a tool—it actively shapes human relationships: The framework proposes that AI doesn't simply facilitate human-to-human interactions but fundamentally mediates them through specific mechanisms, meaning we need to study how AI changes the nature of human exchanges, not just how humans use AI.

Fragmented research communities are missing critical insights: Different academic fields studying human-AI relationships (social computing, machine learning, social sciences) use incompatible perspectives and terminology, preventing researchers from building on each other's findings and slowing progress in understanding responsible AI deployment.

A unified framework could bridge disciplinary divides: By extending Social Exchange Theory to include AI mediation, the proposed AI-MET framework offers a common language that researchers across disciplines can use, potentially accelerating interdisciplinary collaboration and more comprehensive understanding of human-AI systems.

Understanding mediation mechanisms is key to responsibility: The paper identifies that AI influences human relationships through specific mediation mechanisms (though the abstract doesn't detail these), suggesting that responsible AI requires mapping and studying exactly how AI changes human exchanges rather than treating it as a neutral intermediary.

Scale matters in human-AI relationships: The framework explicitly considers both micro-level (individual interactions) and macro-level (systemic) perspectives, indicating that responsible AI governance requires understanding effects across different scales simultaneously.