The project is developing AI-powered approaches to address online hate speech
Dr. Marc-André Kaufhold, Program Area Coordinator (INF) and scientist at the LOEWE Center emergenCITY, as well as a researcher in the PEASEC research group at TU Darmstadt, will take over coordination of the ATHENE project “CYNTRA” alongside Dr. Thea Riebe starting in July 2026. The project focuses on developing an AI-supported ecosystem for analyzing and prioritizing online hate speech. The project aims to help law enforcement agencies and reporting centers evaluate and process such reports more efficiently.
CYNTRA takes a multi-layered approach: The AI models used are designed not only to detect hate speech but also to identify various characteristics simultaneously. For example, the system is designed to identify which groups the content targets or which legal aspects might be relevant. Another key focus is on the continuous review and adaptation of the models. As Dr. Marc-André Kaufhold emphasizes:
“Hate speech is constantly changing-both in terms of language and in the legal definitions of different countries. That is why AI results must remain transparent, verifiable, and adaptable.”
Case studies are also planned in Germany and the United Kingdom to test the developed technologies in real-world scenarios and further refine them for different application contexts. Furthermore, the approaches developed in the project hold potential for other fields of application, such as emergency management and cybersecurity, where digital information must be quickly identified, classified, and made usable for decision-making.
CYNTRA thus addresses a research area that is highly relevant both technologically and socially: the traceable, verifiable, and adaptable use of AI in sensitive, security-relevant contexts.
ATHENE Projekt zu KI gegen Hassrede im Internet: Dr. rer. nat. Marc-André Kaufhold stellt sich vor
