Bauhaus-Universität Weimar Launches €1M 'Lantmark' Project to Fight Digital Manipulation
Bauhaus-Universität Weimar has launched a pioneering research project, 'Lantmark', funded by the Federal Ministry of Education and Research (BMBF) with approximately 1 million euros. Commencing in April 2025, this project seeks to enhance transparency and trustworthiness in digital communication by combating manipulation and identifying automatically generated content.
The 'Lantmark' project, led by Prof. Dr. Benno Stein, focuses on bimodal authentication, merging text and audio analysis. It explores practical procedures to tackle manipulation in digital spaces, with a key emphasis on text watermarking and modifying large language models to carry digital signatures. This innovation could significantly benefit models like ChatGPT, enabling certified versions in the future.
The project, involving Bauhaus University Weimar, Fraunhofer IDMT, and Artefact Germany, aims to make automatically generated content recognisable. As artificial intelligence increasingly writes texts, proving authorship has become a pressing concern. The 'Lantmark' project directly addresses these issues.
With a budget of around 1 million euros from the BMBF, the 'Lantmark' project at Bauhaus-Universität Weimar is poised to make a significant impact on digital communication transparency. By developing bimodal authentication methods and enhancing large language models, the project seeks to strengthen trust in the digital world.
