Abstract

Artificial Intelligence (AI) is transforming organizations enabling novel applications and opportunities. However, the increasing deployment of complex and autonomous AI systems introduces significant challenges in ensuring safe, accountable, and compliant operation. Human oversight has emerged as a critical measure to mitigate AI risks, further emphasized by regulatory mandates like the EU AI Act; yet effective implementation remains underexplored. Employing a socio-technical lens, this research investigates the key challenges organizations face when establishing human oversight mechanisms and identifies essential system and managerial capabilities required for effective human oversight. Drawing on a systematic literature review, we categorize these challenges and IT governance-related capabilities, providing a framework to guide practitioners and researchers in analysing organizational setups and readiness for the introduction of human oversight as a key AI governance mechanism.

Download

Share

COinS