• Priority Area Boosting Fact-checking Activities in Europe
• Funding Round / Action Types 5th Funding Round – Scale-up Projects
• Year 2023, 2024
• Country Netherlands
• Project Status Ongoing

The rise of Large Language Models and other AI-like botnet SaaS has made fact-checking more important than ever. These models can generate human-like text that may contain false information on a large scale and spread it through authentic-looking botnets. This increase in efficiency in generating and spreading factually incorrect information is problematic, as current fact-checking is incredibly time-consuming and done mostly manually. Therefore, this project will automate several parts of the fact-checking process and, in addition, feed the technical and methodological fact-checking methods by in-depth case analysis.

We will develop several AI systems that will find relevant sources that can extract the different claims and cross-reference relevant statements. This automation will greatly increase the efficiency of fact-checking. In addition, there will be in-depth social media case analyses of false claims and related narratives that will expose the origins and dynamics of false information. Knowledge and insights on how and why these false narratives and claims are spread will help improve the fact-checking capability of a news agency both technically and methodologically.

The project team is composed of three parties: xDNA, Tilt and a multimedia news agency fact-check team. xDNA will contribute with their broad expertise in the development of AI systems and develop the automation systems. The fact-check team, will provide relevant domain knowledge and fact-checking expertise. Tilt will bring in expertise and their AI-augmented research environment for online manipulation to provide a better understanding of the origins and dynamics of false information.