TL;DR
The Leiden Declaration on Artificial Intelligence and Mathematics, endorsed by the International Mathematical Union and signed by Fields Medal winner Peter Scholze, calls on mathematicians to confront how AI companies are using published research without consent, circumventing peer review, and threatening the integrity of testing and attribution.
A coalition of mathematicians from institutions including Oxford, Cambridge, ETH Zurich, Columbia and Northwestern has published a formal statement calling on the mathematics community to confront the threats that artificial intelligence poses to their discipline. He Leiden Declaration on Artificial Intelligence and Mathematicspublished on Monday and endorsed by the International Mathematical Union, is the most significant collective response from a major academic discipline to the way AI companies use and, in some cases, exploit published research.
The 11-page document does not oppose AI in mathematics. It opposes the way AI companies are treating mathematical work: training models on published papers without consent, announcing results through press releases instead of peer review, undermining attribution, and reshaping research priorities to serve commercial interests rather than intellectual importance. “Mathematics is, and should always remain, a deeply human activity,” said Ulrike Tillmann, vice president of the IMU.
Five threats to mathematical research
The Statement identifies five specific ways that AI threatens the values that make mathematics trustworthy. First, current AI systems produce plausible but unreliable arguments that are difficult to distinguish from correct evidence. This applies not only to informal reasoning but also to formal computer-coded proofs, where the difficulty lies in the translation between human and machine representations of concepts. The problem of AI-generated content that looks authoritative but contains subtle errors It is not exclusive to mathematics, but in a discipline based on certainty, it is existential.
Second, AI models trained from published mathematical works do not adequately cite the human contributions they synthesize. The Statement notes that much training data was obtained by “systematically exploiting licenses and access agreements that were not made with artificial intelligence in mind, or even simply violating copyright protections.“
Third, the use of AI is incentivizing itself, distorting hiring, funding, and recognition. Fourth, results are increasingly communicated through press releases and blog posts rather than peer-reviewed journals, seeking publicity.within market deadlines before accepted community assessment processes in mathematics can take place.” The Statement cites Google DeepMind’s AlphaProof, which solved three problems from the International Mathematics Olympiad in 2024, but took more than a year to publish its methods in a peer-reviewed venue. Google’s broader AI strategy It relies on mathematical reasoning abilities as evidence of general intelligence, creating commercial incentives to announce results before the mathematical community can properly evaluate them.
Fifth, the autonomy of mathematics is threatened. Research questions may be prioritized because they are amenable to automation and not because they are considered deeply meaningful by experts. “In fact, a broader understanding of the field may be permanently lost in the automation process,”warns the Declaration.
What do you recommend?
The Declaration makes recommendations at four levels. Individual mathematicians must disclose all use of AI tools in their papers, retain personal responsibility for the accuracy of results, refuse to grant authorship to AI systems, and “Carefully consider which tools to use”depending on whether its developers align with the values of the Declaration.
Mathematical organizations should insist that results obtained using automated techniques meet standards that address the specific risks those techniques introduce, protect the rights of authors by developing licensing agreements that prevent the use of published work as training data without consent, and require that results continue to be published through peer-reviewed venues. European regulatory frameworks provide a model, but the Declaration maintains that the mathematics community must also set its own standards independently of the government.
For policymakers, the recommendations are compelling. “Don’t believe the hype“, states the Declaration. “There is currently a strong commercial incentive on the part of the technology industry to exaggerate the capabilities of their products..” It calls for significantly greater public oversight of the AI industry and investment in public computing infrastructure as an alternative to proprietary systems.
who signed it
The Declaration has significant weight due to its signatories. Peter Scholze, winner of the Fields Medal and director of the Max Planck Institute for Mathematics, backed this up with a personal statement: “I am reflecting on my mathematical ideas without the use of AI and generally avoid reading AI-generated texts as best I can..” Other endorsements came from Robbert Dijkgraaf, former Dutch minister of education and president-elect of the International Scientific Council, and Steven Strogatz, Cornell Distinguished Professor for the Public Understanding of Science and Mathematics.
Kevin Buzzard, a professor at Imperial College and one of the most prominent proponents of formalized mathematics, called it “a thoughtful response to what is currently happening, as AI continues to disrupt this space.” The tension between AI capability and research integrity What the Declaration describes is not limited to mathematics, but mathematicians are among the first academic communities to respond with a coordinated and institutionally endorsed declaration.
The deepest argument
The most provocative section of the Declaration addresses AI companies directly. He argues that technology companies are attracted to mathematics because formalized proofs can be verified automatically, creating a “effectively unlimited source of feedback for training artificial intelligence models.The strategy is based on the assumption that capabilities developed through the proving of mathematical theorems will extend to broader general reasoning, an assumption that the Declaration treats with skepticism.
“Some of the resulting general-purpose models are being marketed for applications that raise serious ethical concerns,“, write the authors, “including war, oppression, mass surveillance and the undermining of democracy.” The intersection of AI research and military applications has become one of the defining tensions of 2026, and the Leiden Declaration makes clear that mathematicians do not want their work to be used as training data for systems implemented in those contexts without their consent.
The Declaration was developed over eight months by a 17-member working group following a workshop held in September 2025 at the Lorentz Center in Leiden. It had 37 verified signatories on its first day and is open to additional signatures from the mathematics community.
