Just like humans, AI-based large-language models have characteristics such as morals and values. However, these are not always transparent. Researchers of the University of Mannheim and GESIS – Leibniz Institute for the Social Sciences have now analyzed how the settings of the language models can be ...