07-08-2026

Artificial Intelligence is transforming tax research and practice

Artificial intelligence can already reliably support tax-related work today – but it cannot replace humans: That was a key learning of the 21st annual conference of the working group on Quantitative Tax Theory, which was recently held at the Ingolstadt School of Management. About 60 scientists discussed current research on issues relevant to tax policy.

Organized at the invitation of Prof. Dr. Reinald Koch, Chairholder of Business Administration and Business Taxation, the conference focused on empirical and analytical research in three major areas: the effects of a global minimum tax, the consequences of a carbon tax, and the use of large language models in tax research.

AI is changing the practical field and professional training

The keynote speakers Stefan Groß from the tax and auditing firm PSP in Munich and Dr. Niels Weith from the Institute for Digitalization in Tax Law in Berlin discussed the impact of artificial intelligence on tax practice and tax research with the participants. As an example, the two speakers demonstrated the automated tax classification of VAT chain transactions, in which AI-based solutions currently achieve a reliability rate of about 90 percent. However, Groß and Weith made it clear during the discussion that AI cannot replace humans in this regard: Only the combination of human expertise and AI support leads to better results. As a result, responsibilities of professionals in the field are increasingly shifting – away from preparing solutions and toward a more thorough and intensive review of AI results.

These changes in everyday professional life are also having an impact on tax education at the university level, emphasizes KU tax expert Reinald Koch: "However, a strong professional education remains essential." "Students must be empowered to critically examine AI results and even challenge them – and that’s not possible without a deep understanding of the subject matter." Stefan Groß offered specific suggestions for teaching: He believes a hackathon – where students work on practical solutions to specific problems while receiving targeted support – is a worthwhile endeavor; it is a format he would like to offer himself in the future.

New opportunities for tax research

Tax research itself also benefits from the use of AI: Large language models can thus be used to generate new datasets that contain significantly more detailed information than standard databases. Consolidated financial statements, for example, provide more precise information on the composition of tax expenses, which in turn enables more accurate and nuanced analyses.

During the annual conference, Reinald Koch was elected the new chairperson of the board of the Arbeitskreis Quantitative Steuerlehre e.V. The working group was founded in 2004 to establish a network of researchers engaged in quantitative research. What the members have in common is their economic research into the effects of taxation on the success and behavior of companies and entrepreneurs, using both theoretical modeling and empirical methods.

Text: Eva Schmatloch; translation: Sylvia Schmager