Controlling im Zeitalter der Künstlichen Intelligenz – Grundlagen, Potenziale und Erfolgsfaktoren

Abstract

In today's dynamic and changing business environment, controlling is confronted with numerous challenges. Controllers must simultaneously process increasing amounts of data, make decisions and take on the role of a management partner. Artificial intelligence offers an opportunity to increase the efficiency of controlling and optimize the processing of large amounts of data. In the course of this thesis, the potential of AI in controlling was analyzed through comprehensive literature research before testing its practical applicability through an experiment. The use of AI can bring significant advantages, especially in the area of data preparation and analysis. New technologies can analyze huge amounts of data, process it and recognize correlations that are hidden to humans. This can increase the quality and scope of the data included. An improved database can in turn have a positive impact on other controlling processes such as risk management, planning and reporting. Although AI offers numerous potentials in controlling, its use in companies varies greatly. While the opportunities are generally recognized, there is often hesitation due to a lack of awareness or fear of risks. Challenges arise in the area of data security, due to resistance within the workforce or as a result of inadequate IT-infrastructure. On the other hand, AI can achieve considerable increases in efficiency, enabling controlling to evolve into a management partner. As the literature recognizes a great potential for artificial intelligence in controlling, the actual applicability was tested in an experiment with Microsoft Copilot. The task of the large language model was to carry out a cost center analysis and to create a subsequent management report. However, it was found that the quality of the results is not yet sufficient to carry out a fully automated variance analysis but can serve as support. It also became clear that the results are heavily dependent on the commands used. The results of the work show that artificial intelligence can be used as a tool in controlling and has the potential to improve the degree of automation and the quality of the decisionmaking basis. At the same time, it is clear that its use can only be seen as supportive and that the activities cannot be performed autonomously. In order to utilize the full potential of AI in controlling, sustainable integration must also take place, taking into account aspects such as change management, AI governance and a suitable technological infrastructure.

Date of Award	2025
Original language	German (Austria)
Supervisor	Christoph Eisl (Supervisor)