Legal responsibility for errors caused by artificial intelligence (AI) in the public sector Ahmed Oudah Mohammed Al-Dulaimi Faculty of Law, University of Anbar, Ramadi, Iraq, and Centre de Droit Public Compare (CDPC), Pantheon-Assas University, Paris, France, and Mohammed Abd-Al Wahab Mohammed Faculty of Law, University of Anbar, Ramadi, Iraq Abstract Purpose – This paper aims to assume the responsibility of examining the shifting patterns of legal liability for failures that result from the integration of artificial intelligence (AI) in the public domain. It explores aspects such as the current legal implications, accountability mechanisms of AI errors and potential concerns and issues and proffered solutions for the complex issues that surround AI-related mistakes in public administration. Toward this end, the study outlines a central problem that is defined by the complex nature of errors that arise when AI is applied within the public service. Design/methodology/approach – AI systems have recently been implemented into the public sectors and have influenced positive changes in efficiency and decision-making. However, the development and complication of AI technologies have raised profound worries on accountability in the case of mistakes in public sector. Findings – As international governments increasingly rely on AI for critical selection and planning processes, establishing a clean prison system to educate and allocate responsibility when errors occur is paramount. What it has been found to have the potential to guide policy makers, criminologists and AI planners toward the challenges of implementing AI in the public sector easy to navigate. Finally, the research seeks to assess the potential of AI in public administration and will also serve to create a certain level of transparency, accountability and public trust. Research limitations/implications – To provide a comprehensive response, the research employs a multifaceted methodology that encompasses a thorough literature review, in-depth legal analysis, regulatory assessment, exploration of various liability models, consideration of challenges and ethical considerations and real-world case studies. This holistic approach aims to shed light on the intricate web of legal responsibility and accountability entwined with AI in the public sector. Practical implications – Although as a tool, AI is different from the human agents who use it, and defining and attributing legal responsibility for such errors becomes a challenging task because of the classification of AI as either software or a tool, and the accountability of its human users. Social implications – Consequently, the primary research question emerges: “‘Employing’ AI in the public sector: how can legal responsibility for errors be assigned and governed in ways that respond to the plural employment-aspects of AI?” Originality/value – The significance of this research lies in its ability to address the emerging challenges associated with AI adoption in the public sector. As international governments increasingly rely on AI for critical selection and planning processes, establishing a clean prison system to educate and This research was supported by the University of Anbar/Ministry of Higher Education and Scientific Research, Iraq. Special thanks are extended to Professor Idris Fassassi, director of the Centre de Droit Public Comparé (CDPC) at Université Panthéon-Assas-Paris 2, for graciously hosting this research project during my stay and for his invaluable guidance and support. International Journal of Law and Management Received 29 August 2024 Revised 19 December 2024 Accepted 31 January 2025 International Journal of Law and Management © Emerald Publishing Limited 1754-243X DOI 10.1108/IJLMA-08-2024-0295 The current issue and full text archive of this journal is available on Emerald Insight at: https://www.emerald.com/insight/1754-243X.htm