AI-Enabled and Data-Driven Decision Support Systems for Project Risk Management

Projects are becoming increasingly complex and exposed to multiple sources of uncertainty, making effective risk management essential. At the same time, AI and data-driven technologies are changing the landscape of decision-making across domains. These approaches offer opportunities and challenges for the wider adoption and implementation of predictive and adaptive Project Risk Management (PRM). However, despite the rapid growth of AI applications, their integration into PRM frameworks remains limited and conceptually fragmented. This thesis conducts a Systematic Literature Review (SLR) to provide an overview of existing AI and data-driven applications for PRM and to identify gaps and opportunities for adoption. Fourteen systematic literature reviews between 2003 and 2025 covering domains such as construction, infrastructure, innovation management and small-medium enterprises were collected from Scopus, in accordance with the PRISMA 2020 guidelines. The studies were then compared, thematically coded and categorized based on PRM phases, AI methods, and contexts. The results show an uneven but growing interest in AI-enabled PRM. Machine Learning (ML), Natural Language Processing (NLP), Big Data and Internet of Things (IoT), Building Information Modeling (BIM) and Digital Twins (DT) are the most relevant technologies, mainly applied in risk identification, analysis and monitoring. However, lifecycle coverage, interoperability and ethical governance remain underdeveloped. These findings enable the construction of a multidimensional framework for AI-enabled PRM, which includes technological, processual and contextual dimensions. This framework aims to explain the interplay between smart technologies, risk management processes and organizational facilitators to develop adaptive and transparent decision support systems. The study concludes with a research agenda that seeks to empirically validate and apply the proposed framework across sectors.