A SOCIALLY AWARE AI-AGENT FRAMEWORK FOR ENHANCING TRUST AND ETHICAL INTERACTION IN IOT-MEDIATED SOCIAL ENVIRONMENTS

Abstract

The rapid diffusion of the Internet of Things (IoT) has fundamentally transformed social interaction by embedding intelligent and interconnected devices into everyday human environments. Prior research has extensively explored this transformation through bibliometric and thematic analyses, identifying critical issues related to trust, ethical behavior, and decision-making in IoT-mediated social systems. However, existing studies remain largely descriptive and do not offer practical computational mechanisms to address these social challenges. This paper addresses this limitation by proposing a socially aware AI-agent framework designed to actively enhance trust and ethical interaction within IoT-enabled social environments. The proposed framework integrates social context awareness, trust evaluation, and ethical reasoning to support human-centered decision-making without replacing human agency. A conceptual and simulation-based evaluation is conducted using representative IoT social interaction scenarios, including shared smart spaces and collaborative decision contexts. The results indicate improved consistency in trust-based interactions, reduced ethically conflicting outcomes, and enhanced transparency in mediated decisions. By operationalizing social trust and ethics as computational constructs, the proposed framework bridges the gap between social science theory and practical AI-based intervention. This study contributes a novel interdisciplinary approach that advances IoT research beyond descriptive analysis toward socially responsible and intelligent system design.