A Comparative Analysis of the Role and Requirements of Human Oversight on Artificial Intelligence in EU Law and Iranian Laws

Parsa, Nahid

doi:10.22034/jrpl.2025.721655

A Comparative Analysis of the Role and Requirements of Human Oversight on Artificial Intelligence in EU Law and Iranian Laws

Document Type : Original Article

Author

Nahid Parsa

PhD in Private Law, University of Mazandaran, Mazandaran, Iran

10.22034/jrpl.2025.721655

Abstract

The AI Act 2024 introduced by the European Commission highlights human oversight as a key safeguard for AI systems, setting a comprehensive legal framework for monitoring AI operations within the European Union. This legislation establishes clear requirements for the supervision of AI systems, particularly those categorized as high-risk. In contrast, Iranian law currently lacks specific regulations addressing human oversight in AI applications. However, indirect references to oversight responsibilities can be found in legal provisions such as Articles 7 and 12 of the Civil Liability Law, Article 333 of the Civil Code, Articles 69 and 112 of the Maritime Law, and Article 15 of the Mandatory Insurance Law of 2016. This article explores the nature of human oversight obligations in both legal systems by addressing what should be supervised, when oversight is required, and who bears the responsibility for supervision. It further highlights the uncertainties and gaps in the AI Act 2024 and the potential risks of excessive reliance on providers for ensuring the safety infrastructure of high-risk AI systems.

Keywords

human-centric

human oversight

EU AI Act

automated decision-making