Leiden Declaration on AI and Mathematics
How can artificial intelligence support mathematical research without weakening transparency, recognition of human work and scientific responsibility? This question is at the heart of the Leiden Declaration on Artificial Intelligence and Mathematics, published on 2 June 2026. An international working group developed it following a conference at the Lorentz Center in Leiden. The International Mathematical Union (IMU) has endorsed the declaration.
The declaration recognises opportunities for new discoveries but warns of plausible yet incorrect proofs, inadequate attribution and premature public claims of breakthroughs. Among its recommendations are disclosing the use of AI and other tools, keeping results independently verifiable, and leaving responsibility for correctness and attribution with human authors. Its recommendations also address mathematical organisations, funders, policymakers and AI companies.
Signatories from Münster: The list includes the following Mathematics Münster researchers: Arthur Bartels, Christopher Deninger, Siegfried Echterhoff, Konstantinos Kartas, Thomas Nikolaus, Peter Schneider, Eduardo Silva, Sam Shepherd, Edoardo Giovanni Tolotti, Alexander Van Werde, Floris Vermeulen and Yifei Zhao. Their signatures are personal statements; an institutional affiliation does not imply an endorsement by the university.
A further discussion from Münster: Thomas Nikolaus launched the independent AI.MATH initiative together with Claudia Alfes in Bielefeld and Andreas Thom in Dresden. It offers draft positions on the future of mathematical research and degree programmes in the age of AI, inviting questions and disagreement. It provides another forum for the mathematical community to discuss issues also raised by the Leiden Declaration.