As a form of Artificial Intelligence, large language models, are able to produce texts whose linguistic level corresponds to that of an educated human, or to communicate with humans (e.g. through a chatbot). The judicial system requires the production, interpretation, specification and application of legally relevant texts. Some players within the judicial system are already using learning language systems such as ChatGPT for support. However, such activities are largely uncoordinated and unsystematic. The utilization of learning language systems could help make the judicial system more effective and efficient in fulfilling its functions, thus strengthening fundamental rights, democracy and the rule of law.

The key question of the Research Network should be the question of which challenges the support by learning language systems will bring for the acceptance and acceptability of the judicial system, and how to deal with them. The interdisciplinary research consortium, which ideally should bring together relevant transdisciplinary research perspectives, should forecast, analyze and evaluate potential scenarios for the application of learning language systems in the judicial system, and make suggestions on how to design the applications and the transformation processes.

The foundation is providing a sum of up to 1.5 million euros for a period of three years. Under the leadership of a scientific coordinator who acts a spokesperson, a research topic is investigated by several working groups at various scientific institutions in Germany and abroad. The members of a research network are granted a high degree of freedom, determine their research programs independently, and have the opportunity to meet regularly in Ladenburg on invitation by the Foundation.

© Daimler und Benz Stiftung - ArtemisDiana