Date: Tuesday, January 27, 2026 at 8:00 am - 9:00 am ET
Leadership Role of Bayesian Statistics in AI Webinar
Dr. Nadja Klein, a Professor in Statistics and Leading the Research Group Methods for Big Data at Karlsruhe Institute of Technology, Germany, introduced the COPSS-NISS Leadership Webinar, highlighting the collaboration between COPPS and NISS to promote leadership skills in statistics. The webinar's theme explored the leadership role of Bayesian statistics in an age of AI, and exploring how Bayesian thinking can make AI more transparent, trustworthy, and impactful. The webinar featured presentations by David Dunson and Xuming He. Panelists David Dunson and Xuming He were introduced, with their extensive research and leadership experience in Bayesian statistics and data science. The session was structured to include each panelist's leadership journey, a discussion moderated by Nadja on how leadership roles in Bayesian statistics have changed and should be changed, what roles and contributions Bayesian statistics can make to AI, and recommendations on successful interdisciplinary collaborations and the next generation. At the end, an audience Q&A session was opened.
Bayesian Statistics in AI Research with David Dunson
David discussed the role and value of Bayesian statistics in the age of AI, emphasizing its potential to enhance AI methodologies and lead scientific research. He highlighted five key areas where Bayesian statisticians can contribute, including improving theoretical understanding of AI models, addressing robust and reproducible inferences, handling small data settings, dealing with selection bias, and modeling AI-derived data. David stressed the importance of statisticians, particularly Bayesians, taking active leadership roles in AI research to ensure scientific soundness, ethical responsibility, and societal benefit.
Bayesian Methods in AI Era with Xuming He
Xuming shared his perspective on Bayesian thinking and Bayesian methods, emphasizing their practical applications and computational feasibility. He discussed the integration of data, experience, and domain knowledge, which is becoming increasingly important in AI and multi-source data analysis. Xuming also highlighted the need for Bayesian statistics in synthetic data generation and the analysis of rare events, as well as the opportunity for everyone to contribute to the design, evaluation, and impactful applications of Bayesian methods in the AI era. The panelists discussed the importance of statisticians taking leadership roles in AI initiatives, engaging with computer scientists and domain experts, and promoting their work to gain trust and recognition. They emphasized the need for statisticians to be proactive, learn AI terminology, and contribute to interdisciplinary research. The conversation ended with a discussion on uncertainty quantification in complex Bayesian models and the need for further research in this area.
Thanks & Acknowledgement
We extend our sincere thanks to our distinguished speakers, David Dunson and Xuming He, for their thoughtful and inspiring presentations, and to our moderator Dr. Nadja Klein for her leadership and for guiding a rich and engaging discussion. We are grateful to the organizing committee for their dedication in making this event possible, to COPSS for their continued collaboration in advancing leadership in statistics. Finally, we thank all participants for their insightful questions and for contributing to a meaningful conversation on the role of Bayesian statistics in the age of AI.