They fear that scientific research will become beholden to commercial tech giants, that trade journals will be flooded with unreliable results, and that scientists' autonomy will erode. They call on colleagues to take action and make recommendations in the Leiden Declaration on Artificial Intelligence and Mathematics. Other scientists are invited to sign the document.
The initiative grew out of a workshop at Leiden University last year. A team of sixteen international researchers, convened by TU/e mathematician Jim Portegies , then drew up the final statement.
The group formulated five main risks of AI for mathematics: unreliable results, missing citations, dependence on closed commercial systems to contribute to mathematical research, exaggerated claims and the loss of scientific independence.
Among the signatories are former Dutch Minister of Education Robbert Dijkgraaf and Terence Tao, the most important mathematicians alive today.
What happens when a mathematical proof is no longer the work of a human, but of a closed AI model to which academics have no access, the creators of the declaration ask. Who is responsible for any errors, and who gets the credit if the argument is correct? How can one determine whether an AI-generated result is truly new, or just a clever restatement of existing work without proper attribution?
"Furthermore, the rapid rise of AI is dragging the underlying mathematical models straight into the public debate", Portegies elaborates. "Technologies developed on this basis directly intersect with sensitive ethical issues, ranging from military applications and mass surveillance to political disinformation and environmental damage."
Not a ban, but recommendations for responsible use
The researchers do not want to ban AI, but they do want to spur their colleagues to take action and make recommendations for individuals and organizations, such as openness about AI use, careful acknowledgment of sources, and ensuring that a human always has final responsibility. They also want additional policy from universities, scientific journals and subsidy providers that protects the quality of mathematical research. In addition, politicians should regulate the AI sector more strictly and invest in public alternatives, so that knowledge and power do not lie entirely with commercial tech companies.
Top international mathematician Terence Tao signs too
The Australian-American scientist Terence Tao has also signed the Leiden Declaration. He is widely regarded as one of the most important mathematicians working today. For his research, he received the Fields Medal, the highest honor in mathematics. Although Tao works extensively with AI himself, using it to check his own proofs, he stated upon signing: "This has been the result of months of community input about the fundamental values and goals of the mathematical community. In retrospect, these were questions we should have been systematically discussing years ago, but in any event the exercise was extremely valuable, and the end result is excellent. I wholeheartedly endorse the statements and recommendations in this declaration."
Photo: UCLA
Recently, an internal model from tech company OpenAI (maker of ChatGPT) succeeded in solving a well-known mathematical problem (the so-called unit distance problem , which concerns the placement of points in a plane, in such a way that as many points as possible have distance one to each other). Although the mathematical proof has been shared, the model used to obtain the result and the way it was trained remains a closed book.
"This case raises major concerns about closed access, insufficient attribution of related ideas, and lack of transparency about methods and the use of sources" says Rodrigo Ochigame from Leiden University, who co-organized the workshop where the idea for the declaration was born.
'Mathematics should remain a human endeavour'
The International Mathematical Union (IMU), the worldwide umbrella organization of mathematicians, endorses the declaration . "We take the rapid development and impact of artificial intelligence on our field extremely seriously," says Ulrike Tillmann, Vice President of the IMU.
'It opens new and exciting opportunities, but it also raises questions that cannot be left unexamined. By endorsing the declaration, the IMU affirms that the future of mathematical research must be guided by human judgment, fair and transparent practices, and the shared values of the global mathematical community. Mathematics is, and should always remain, a profoundly human endeavour.'
The Leiden Declaration on Artificial Intelligence and Mathematics is an initiative of individual researchers. Other scientists can sign the declaration themselves. On 26 July 2026, Jim Portegies will present the Leiden Declaration at the International Congress of Mathematicians (ICM) in Philadelphia.
