The manufacturing industry is becoming increasingly complex, dynamic, and connected. As a result, companies are facing challenges in managing highly nonlinear and stochastic activities due to the many uncertainties and interdependencies they face. In recent years, the development of artificial intelligence (AI) has shown potential for transforming the manufacturing domain through the use of advanced analytics tools for processing large amounts of manufacturing data. There are many examples of research investigating how AI can be used to optimize performance in manufacturing companies. However, studies show that only a small percentage of firms across industries engage in widespread adoption of AI. Most companies only run ad hoc pilots or apply AI to a single business process. Some researchers claim that AI has the potential to disrupt the manufacturing industry as we know it, arguing that companies that only partially commit to an AI transformation will be outcompeted by those who can offer new data-driven services. Others, however, question the current direction and potential of AI. This thesis investigates, from a multidisciplinary point of view, what is required from a manufacturing company leader to efficiently support a widespread introduction of AI. It combines quantitative and qualitative methods combined with real-life examples of introducing AI in an aerospace manufacturing company. While the term AI is commonly used in various contexts, there is some confusion surrounding the concept, and a vast spectrum of definitions have emerged. Therefore, the thesis begins with describing essential concepts within the field of AI as well as their history. Leaders of manufacturing companies play an important role in achieving widespread implementation. However, there is limited research on how leaders best can contribute to widespread implementation. The first step of this research was therefore to create a capability framework for leaders of manufacturing companies that wish to introduce AI on a wide scale. In this work, we identified a willingness to learn about AI as one of the most important capabilities leaders could benefit from when supporting the widespread adoption of AI. However, this is challenging. The field grows as technologies emerge that could fit under the AI umbrella. Additionally, researchers and practitioners lack a coherent definition of AI, which has led to a mystification of the term. 9 To gain an understanding of the leadership capabilities required for an effective AI implementation, it is necessary to consider the nature of AI technology itself and its use. Technological advancements within AI have enabled non-experts to develop AI applications that previously required in-depth computer science, statistics, and mathematics knowledge. To develop the understanding of these technological advancements, we conducted a second study that specifically examined the role of abstraction in AI. Our aim was to quantify the level of abstraction in deep learning by investigating the number of lines of code utilized in deep learning projects. We saw a dramatic decrease in the number of lines of code used, indicating an increase in abstraction. Thereafter, we proceeded to investigate implications of this abstraction increase, particularly with respect to mimesis. While these developments contribute to the democratization of AI, promoting collaboration and allowing us to do more with less, they also come with several drawbacks such as convergence on suboptimal solutions, quality assurance problems and scarcity of talent that most organizations are not yet strategically prepared to handle. Much of the research on AI in manufacturing has been conducted outside the daily operations of manufacturing companies. Further research has been requested on introducing it into daily operations. Against this background, in order to investigate implications, considerations, and trade-offs that need to be made when introducing AI into daily operations I conducted an action research study using a case study of a global manufacturing company deploying AI to develop capabilities and enhance decisionmaking. This research offers a multidisciplinary investigation of some of the challenges and considerations involved in implementing AI in the manufacturing industry and highlights the role of manufacturing company leaders in facilitating the adoption of AI in the industry. While AI can be a useful tool for solving problems, it may not always be the best solution, and companies should carefully consider where and how to use it. The research clearly shows the importance of multidisciplinary approach, collaboration, as well as the combination of competencies to succeed in AI transformation.