On July 1, 2026, a delegation from the Institute for AI International Governance (I-AIIG) at Tsinghua University visited the Berlin Institute for the Foundations of Learning and Data (BIFOLD) to exchange views with researchers from BIFOLD and Charité – Universitätsmedizin Berlin on trustworthy AI, explainable AI, frontier AI safety evaluation, and international AI governance. The I-AIIG delegation included Research Fellows Wang Jingyu, Sheng Shuyang, and Song Yuxin, along with Chen Fan, Foreign Affairs Assistant.
As one of Germany’s national AI competence centers, BIFOLD is jointly hosted by Technische Universität Berlin and Charité – Universitätsmedizin Berlin. The meeting, hosted by Professor Dr. Konrad Rieck, BIFOLD’s Deputy Director, brought together leading researchers, including Professor Wojciech Samek, Professor Grégoire Montavon, and Dr. Elke Witt.
During the meeting, the I-AIIG delegation highlighted the institute’s recent efforts in global AI and digital governance, as well as international cooperation. The delegation also introduced the China AI Safety & Development Association (CnAISDA), a research network for advancing international scientific collaboration on AI development and safety. In addition, the delegation shared updates on Tsinghua’s AI Agent Hospital initiative, which explores trustworthy AI applications in healthcare across hospitals, community care, and home-based settings.
The delegation also brought up I-AIIG’s ongoing international engagement, including participation in the first UN Global Dialogue on AI Governance, an upcoming official session at the World Summit on the Information Society (WSIS) on AI governance for automated driving systems, and preparations for the World Artificial Intelligence Conference (WAIC) and the International AI Cooperation and Governance Forum 2026. BIFOLD researchers were invited to participate in future dialogue and exchange events.
Researchers from BIFOLD presented their findings on AI security, explainable AI, and AI for healthcare. Professor Rieck shared recent advances in AI security research, including challenges to the robustness of AI systems. Professor Samek also introduced his team’s work on explainable AI and its use in medical and scientific research, highlighting the role of transparency and interpretability in building trustworthy AI. Meanwhile, Professor Montavon discussed the application of explainable AI in clinical settings, emphasizing the importance of interpretability in supporting medical decision-making and improving trust in AI-assisted diagnosis.
The meeting concluded with an in-depth discussion of AI safety research in China and Germany, frontier AI evaluation, explainable AI, medical AI applications, and international cooperation in AI governance. The exchange offered both sides an opportunity to deepen mutual understanding and explore future collaboration across AI research, governance, and international academic exchange.
