1 Apr 2025
Clear problem definition, precisely adapted algorithms, good data: Prof. Dr. Stefan Feuerriegel explains how artificial intelligence can be used successfully in management.
© Florian Generotzky/LMU
Trustworthy AI: people at the center
Prof. Dr. Stefan Feuerriegel, Head of the Institute for Artificial Intelligence in Management at the LMU Munich School of Management, emphasizes: “My job is often to dampen expectations of AI - then we can work productively with it.” AI is not a panacea, but a tool that supports people in making decisions - not replaces them.
His team develops AI models that help companies, politicians and scientists to make well-founded, comprehensible decisions. The key prerequisite: decision-makers remain in the “driver's seat” in all phases of implementation. For example, Feuerriegel's models analyze the dynamics of hate speech in social media, identify fake news or help to allocate funds for development aid in a more targeted manner.
The institute relies on two approaches: interpretable models (e.g. linear regression) and post-hoc explanation techniques (e.g. heat maps for neural networks). In this way, AI remains comprehensible - even with complex algorithms.
AI is ultimately “nothing more than a bit of mathematics, combined with computing time and good data”, says Feuerriegel. His approach: minimally invasive AI that supplements rather than replaces expertise - for example in quality management with digital twins or in data-based decision preparation using counterfactual simulations.
“We show companies what happens when decisions are made - and what happens when they are not,” says Feuerriegel. His claim: tailor-made algorithms for clearly defined problems. Because: “There is no one AI for all cases.”
You can read the entire interview in the newsroom at lmu.de/en.