As of 2026, Fraunhofer IAIS performs research on
generative AI (AI-based curation and synthesis of high-quality training data, improved reasoning skills, knowledge distillation, domain-specific adaptation, multi-agent systems and multimodality),
agentic AI (specialized reasoning, edge AI agents and sovereign AI agent system, reliable agent evaluation, enterprise MLOps, and extended agent capabilities) and
hybrid AI, i.e., the combination of different methods of AI. In its research, Fraunhofer IAIS cooperates with other Fraunhofer organizations (e.g., the Fraunhofer Big Data and AI Alliance, the Fraunhofer Competence Network on Quantum Computing, Fraunhofer CCIT or Fraunhofer HNFIZ) and the universities of
Bonn and
Siegen, leads the regional AI competence platform KI.NRW, and supports the
Lamarr Institute for Machine Learning and Artificial Intelligence. Key research projects include: • OpenGPT-X: development of the European multilingual language model Teuken 7B • OpenEuroLLM: development of an open-source multilingual language model • DeployAI: development of a technology platform for small and medium-sized businesses and the public sector in Europe • Jupiter AI Factory: supercomputer-based development of secure and data protection-compliant AI applications • TrustLLM: development of an open, trustworthy and fact-based large language model for Germanic languages == References ==