When Artificial Intelligence explains its decisions, user trust grows. They gain a better understanding of the environment and shop more consciously. A new study from the University of Liechtenstein, published in the renowned "International Journal of Information Management," demonstrates this.

AI systems often deliver ready-made answers, but their decision-making process mostly remains in the dark. The technology acts as an opaque "black box". Dr. Pejman Ebrahimi and Prof. Dr. Johannes Schneider from the Artificial Intelligence and Data Science research area at the University of Liechtenstein set out to change this. Together with Prof. Dr. Stefan Hoffmann from Kiel University, they developed a chatbot that operates like a glass box.

The researchers relied on the "Chain-of-Thought" method. The chatbot does not present blunt purchase recommendations; instead, it outlines step by step why a product is sustainable. It weighs arguments and makes its logic comprehensible.

The team tested the system with 417 people – and the result is clear: transparent explanations build trust. Users gain a better grasp of the often complex connections within climate protection. Granted, longer explanatory texts require more mental effort and raise privacy concerns when the AI analyzes personal preferences. Yet the benefits outweigh the drawbacks: those who understand the reasons behind a recommendation plan more resolutely to shop in an environmentally friendly way.

To achieve this result, Ebrahimi and Schneider trained their system using modern methods such as LoRA and GRPO. Afterwards, they analyzed the test subjects' behavior in detail.

The study impressively demonstrates how responsible digital technology can pave the way to a more sustainable future. The University of Liechtenstein warmly congratulates the researchers on this outstanding success.