Artificial Intelligence in Organizations: Governance, Control, and Opacity from a Systems-Theoretical Perspective
How can organisations meaningfully govern AI systems when their inner workings are often difficult to understand? This seminar addresses the question of how artificial intelligence (AI) can be designed and implemented responsibly in organisations when faced with uncertainty and opacity. The analysis focuses on governance structures, decision-making processes and control mechanisms for algorithmic systems.
Drawing on cybernetics and systems theory — from Ashby's concept of control and second-order observation to Luhmann's theory of organisations — the seminar combines theoretical foundations with case studies from business, public administration and the military. Rather than treating ethical questions as pre-given principles, they are analysed as practices of observation and attribution within organisational decision-making. Particular attention is paid to how risks, biases and opacity are observed, assessed and processed within organisations.
The seminar aims to develop a deeper understanding of how organisations deal with the 'black box' of AI, and demonstrate that the key issue is not how AI can be made fully transparent, but how organisations can continue to make decisions and take action despite persistent opacity.
Objectives & Competences
Knowledge level: Students are familiar with the key concepts of cybernetics and sociological systems theory, such as first- and second-order observation, steering and control, decision premises and uncertainty absorption, and can relate these to the organisational use of AI. They can situate the problem of AI opacity ('black box') theoretically and distinguish it from a purely technical understanding of transparency. They are familiar with key regulatory approaches, such as the EU AI Act, and their organisational implications.
Skills level: Students can analyse and compare governance structures, attributions of responsibility and forms of control when dealing with AI in different organisational contexts (business, public administration and the military). They reflect on the ethical requirements of AI, such as fairness, responsibility and accountability, as organisational practices of translation and attribution. They can apply systems-theoretical concepts independently to case studies and defend their analyses.