Künstliche Intelligenz in der Lieferkette: Qualitative Untersuchung zur Umsetzung der CSDDD

Abstract

In the context of increasing legal regulations—such as the German Supply Chain Due Diligence Act and the EU Corporate Sustainability Due Diligence Directive—companies are facing growing pressure to systematically identify, assess, and mitigate human rights and environmental risks along their supply chains. Against the backdrop of rising data volumes and increasingly complex global value networks, the question arises as to what extent digital, and particularly AI-powered, information systems can effectively support companies in meeting these regulatory requirements. The objective of this thesis is to explore the potentials, challenges, and areas of application of AIbased tools in the context of corporate due diligence obligations. The study combines a structured literature review with an empirical section consisting of six qualitative case studies across different industries. The theoretical part introduces key international frameworks and voluntary standards, followed by the corresponding legal regulations such as the German Supply Chain Due Diligence Act and the EU Directive on Corporate Sustainability Due Diligence. This is followed by an explanation of the technological foundations of artificial intelligence and an analysis of the use of digital solutions in sustainability management. The focus lies on specific software tools such as Prewave, Integrity Next, Osapiens, Tacto, Retraced, and EcoVadis. The empirical section examines six company case studies with regard to organizational structures, tools in use, and their contribution to the implementation of due diligence obligations. The findings indicate that AI-based systems can offer valuable support, particularly in the areas of risk identification, as well as in increasing transparency and efficiency within supply chains. At the same time, challenges remain, particularly regarding data quality, internal responsibilities, and data protection. The implementation of effective measures based on AI-generated insights still requires human oversight and evaluation. This thesis provides practical insights into the use of AI-supported tools for the implementation of regulatory sustainability requirements and addresses a research gap in a relatively unexplored field. It lays a solid foundation for further research, especially for future quantitative studies on the effectiveness of AI systems in sustainability contexts. Furthermore, it offers practical recommendations for companies aiming to implement or improve AI-based information systems.

Date of Award	2025
Original language	German (Austria)
Supervisor	Andreas Mladenow (Supervisor)