AMS :: Notices of the American Mathematical Society

Notice: AMS websites will be down for maintenance from 6am until 3pm EST on Saturday, August 15.

-->

PDFLINK

A Call for Action: The “Leiden Declaration on Artificial Intelligence and Mathematics”<br>Siobhan Roberts

Introduction

The idea for the “Leiden Declaration on Artificial Intelligence and Mathematics”—published on June 2, 2026—first took shape one year before over pints at a pub near Cambridge University. It was during a workshop at the Isaac Newton Institute for Mathematical Sciences on “Big Proof,” addressing complications of bringing automated proof technology into mainstream math research.

Dame Ursula Martin, a mathematician and computer scientist at Oxford, was joined at the pub by mathematician Michael Harris of Columbia University, and Rodrigo Ochigame, an historian and anthropologist of computing and artificial intelligence at Leiden University. All three were planning on attending a related meeting that September—a workshop on “Mechanization and Mathematical Research” at Leiden University (Ochigame was an organizer). Ultimately, their pub talk landed on: ‘Hey, why don’t we think of something useful that can come out of the Leiden meeting?’

Harris, author of the Substack newsletter “Silicon Reckoner,” arrived in Leiden with a four-point plan for a declaration. He pitched it to the 60 participants, which included mostly mathematicians, but also computer scientists, engineers, historians, philosophers, social scientists, and policymakers. Part of the workshop program was organized around breakout groups; the two breakout sessions on the declaration were oversubscribed.

Subsequently, a working group of 16 authors, in consultation with more than 30 mathematicians independent of the workshop, drafted and redrafted the statement in a process that was at times contentious—the writing and rewriting went on for eight months. Jim Portegies, an applied mathematician at the Eindhoven University of Technology, served as convenor, untangling competing opinions and building consensus.

Figure 1. Rodrigo Ochigame and Mateja Jamnik (front row, center), two coauthors of the declaration, and other mathematicians in September 2025 at a workshop at Leiden University, where the “Leiden Declaration on Artificial Intelligence and Mathematics” initially took shape.

“Artificial intelligence places core values of the mathematical community under threat,” Portegies said in an interview. “Admittedly, the technological developments could advance mathematics—I hope even to the extent that it benefits much more than the field itself. But if we don’t act and organize ourselves, there’s a large risk that AI will do serious harm to the discipline and beyond. This is why the declaration calls for action, with recommendations for individual mathematicians, mathematical organizations and governments and ends with a request to commercial AI to respect the values outlined in the declaration.”

Bryna Kra, a mathematician at Northwestern University and a former president of the American Mathematical Society, was also among the working-group authors. “AI is already transforming our profession: accelerating the pace of research, changing the meaning of authorship, altering the ways students learn, and widening gaps in access to powerful tools,” Kra said in an interview. “Mathematicians must be part of shaping these changes rather than merely reacting to them. This declaration is a call for the mathematics community to engage and develop shared standards and guidelines.”

Upon its publication, the Leiden Declaration was endorsed by the International Mathematical Union. It is open for signing and endorsement by individuals and organizations.

Two Fields Medalists—with divergent views on AI—were early supporters and endorsers:

Peter Scholze, a mathematician at the University of Bonn and 2018 Fields Medal winner commented: “This is a wonderful declaration, coming at the right time. The goal of mathematical research is human understanding of mathematics, and so mathematics can only thrive in a community of human mathematicians. It is crucial to preserve this communal spirit. In my experience, mathematical ideas, like children, must be nurtured and grow over the years. Just like I do not want my children to be educated by AI, I am pondering my mathematical ideas without use of AI, and generally avoid reading AI-generated text as best as I can.”

Terence Tao, mathematician at the University of California, Los Angeles, and 2006 Fields Medal winner, commented: “AI is a powerful technology that shows significant promise for being transformatively useful in mathematics in particular (due in large part to our ability to filter out AI’s mistakes to an extent that is not possible in other disciplines), and yet it also carries great risks and harms if used irresponsibly, or if that usage is dictated by external entities not inherently aligned with scientific goals. The apparent...