THE ETHICAL, LEGAL, AND SOCIAL IMPLICATIONS OF DEPLOYING AGENTIC AI: EXAMINING AUTONOMY, ACCOUNTABILITY, AND HUMAN OVERSIGHT IN HIGHLY AUTOMATED DECISION-MAKING SYSTEMS

The deployment of agentic AI—autonomous systems capable of independent decision-making—raises significant ethical, legal, and social challenges, particularly in relation to autonomy, accountability, and human oversight. This study adopts a mixed-methods approach, integrating a scoping review of 25 scholarly sources published between 2016 and 2024, an analysis of 150 documented AI-related incidents from publicly accessible databases reported between 2020 and 2024, and survey responses from 500 stakeholders engaged in AI governance and policy discourse. The findings indicate that approximately 78% of reported incidents are associated with insufficient human oversight, contributing to accountability gaps in high-risk domains such as healthcare and finance. Emerging regulatory frameworks, including the early provisions of the EU AI Act (2024), emphasize the necessity of human oversight, yet preliminary analyses suggest limitations in operational clarity and enforcement preparedness. Furthermore, survey data reveal that 62% of respondents express distrust toward highly autonomous AI systems, primarily due to perceived risks associated with diminished human control. The study underscores the importance of hybrid human–AI decision-making models to reconcile efficiency with ethical responsibility. It concludes by advocating for interdisciplinary governance strategies that enhance transparency, accountability, and equity, thereby supporting the sustainable and responsible integration of agentic AI into socio-technical systems.