Leiden Declaration on Artificial Intelligence and Mathematics

Signatories

The primary signatory list now lives on the homepage below the declaration text.

Sign Declaration
3134 Signatories
  1. Hongjian Yang ORCID Ph.D., Stanford University
  2. Andrej Srakar ORCID Artificial Intelligence, Jožef Stefan Institute
  3. Paolo Stefano Giudici ORCID Professor of Statistics, University of Pavia
  4. Danilo Gligoroski ORCID Norwegian University of Science and Technology (NTNU)
  5. Chandan Singh Dalawat ORCID Visiting Professor, Ashoka University
  6. Ang Li verified email Guangdong University of Technology
    Comment

    I fully endorse and support the Leiden Declaration and its call to safeguard the integrity, trustworthiness, and human-centered values of mathematical research in the age of artificial intelligence. The Leiden Declaration draws a boundary that mathematics' wisdom cannot be reduced to a format — not by rejecting technology, but by refusing to let mathematical research be degraded into output optimizable by models. The deepest insight of this declaration lies in its sober diagnosis of what truly threatens AI-driven mathematics today: not that machines are becoming smarter, but that trustworthiness itself is collapsing — for when generated proofs can no longer be understood, audited, or attributed by humans, mathematics forfeits that testable dignity which defines it as public knowledge. In essence, it stands as a civilizational constitution for the artificial — reminding us that however technologies may evolve, the duty of proof, the ethics of attribution, and the reverence for understanding itself, remain forever that human craft which no act of cognition may ever outsource.

  7. shangjun shi ORCID East China Normal University
  8. Ankan Man ICTS-TIFR
    Comment

    AI can only be a tool and must require proper supervision. AI can't be trusted.

  9. Martin Ruskov ORCID Ricercatori, Universita' degli Studi di MILANO
    Comment

    Research in AI could only be sustainable if models are open, not only weights, bit also data and tools, as in the Model Openness Framework.

  10. f alberto grunbaum ORCID UC Berkeley
  11. Kyle Ormsby ORCID Reed College Department of Mathematics & Statistics
  12. Max Petschack verified email University of Melbourne
  13. A Sankaranarayanan ORCID Vivo Bio Tech Ltd
  14. Prof. Dan Wu ORCID School of Electrical and Electronic Engineering, Huazhong University of Science and Technology
  15. Chenxing Qian ORCID PhD, University of Pittsburgh
  16. George Turcas ORCID Lecturer, Babes-Bolyai University
  17. Gabor Szekelyhidi ORCID Professor, Northwestern University
  18. Jordan Martino verified email Northeastern University
  19. Maisha Rumman Member of Energytech Cypher
    Comment

    I support the Leiden Declaration

  20. Sam Wang verified email New York University
  21. Jinyang Zhang verified email University of California, Berkeley
  22. Mirza Mehmedagic verified email University of Chicago
  23. Arman Valaquenta ORCID Founder and CEO, Dauðalogn Holdings Limited
    Comment

    We used to take square roots by hand. Then we used calculators. We must remain in charge what to take square roots of now, with AI. Insight is ours.

  24. Benjamin Galluzzo ORCID Executive Director, Consortium for Mathematics and its Applications (COMAP)
  25. Yunchu Dai ORCID Mathematics, Massachusetts Institute of Technology
  26. Brian Chao ORCID Graduate student, Cornell University
  27. Michael A. Boss ORCID PhD, Physics
  28. Lejian Wang verified email Jilin University
  29. Ian Zemke ORCID University of Oregon
  30. Patrick Shafto verified email Rutgers University
  31. Allen Nikora verified email Jet Propulsion Laboratory (retired)
  32. Florian Richoux ORCID Senior researcher, National Institute of Advanced Industrial Science and Technology (AIST)
    Comment

    The Leiden Declaration is a necessary statement regarding the rise of generative AI in scientific research. In fact, most of its recommendations extend beyond mathematics and remain relevant to most of scientific fields.

  33. Peijie Li ORCID The University of Hong Kong
    Comment

    AI brings great productivity to mathematical research, allowing people to focus more on innovative ideas than technical details. It shall be seen as the crystallization of the wisdom of all humanity and shall belong to all the people, be used for the people and by the people.

  34. Rohitesh Pradhan ORCID Ewing Christian College, Prayagraj
  35. He Xin ORCID
  36. Dmitrii Pasechnik ORCID Research Professor, Northwestern University
  37. Alexander Mundey ORCID Research Fellow, Adelaide University
  38. Sang-hyun Kim ORCID Professor, Korea Institute for Advanced Study
  39. Xingkai Wang verified email Pennsylvania State University
  40. Songhua He ORCID Research Assistant, Rutgers, The State University of New Jersey
  41. 富可 王 ORCID National University of Defense Technology
  42. Daxin Xu ORCID Academy of Mathematics and Systems Science
  43. James Propp ORCID UMass Lowell
    Comment

    If you are an individual mathematician, be aware that use of these new tools carries along with it the risk of atrophy of old skills and habits. As you use AI, pay attention to its effect on you.

  44. Zhe Xu ORCID University of Oregon
  45. Shaoyun Bai ORCID Assistant Professor, Massachusetts Institute of Technology
  46. Ada Chan ORCID York University
  47. Rommel Real ORCID Assistant Professor, University of the Philippines Mindanao
  48. Fidel Nemenzo University of the Philippines
  49. Jerry Gong verified email Johns Hopkins University
  50. Sheldon Axler ORCID Professor Emeritus, San Francisco State University
  51. Yu Wang ORCID Assistant Professor, Southwest Jiaotong University
  52. Kshitij Anand Patil verified email Simon Fraser University
  53. Elena Pavelescu ORCID Associate Professor, University of South Alabama
  54. Yukun Cai verified email Pennsylvania State University
  55. Sara Kalisnik Hintz ORCID Associate Professor, Pennsylvania State University
  56. Dimitri Mihaylov verified email University of Arizona
  57. Ruoxi Li ORCID Ph. D, University of Pittsburgh
  58. Sam Hopkins ORCID Associate Professor, Howard University
  59. Vincent Vatter ORCID Professor, University of Florida
  60. Marco Antonio Piedra Venegas verified email Universidad de Costa Rica
  61. Nicholas Beaton ORCID University of Melbourne
  62. Jiwoon Sim ORCID Graduate student, University of Alberta
  63. Blanka Horvath verified email University of Oxford
  64. Mihai Caragiu verified email Ohio Northern University
  65. Louis-Pierre Arguin ORCID City University of New York and University of Oxford
  66. Jyun-Ao Lin ORCID Assistant Professor, National Taipei University of Technology
  67. Changwei Zhou verified email State University of New York at Binghamton
  68. Cathy Li verified email University of Edinburgh
  69. Jiahe Zhang verified email New York University
  70. Sonja Mapes ORCID Northwestern University
  71. Yash Uday Deshmukh ORCID Member, Institute for Advanced Study
  72. T. Kyle Petersen ORCID Professor, DePaul University
  73. Quanlin Chen verified email Princeton University
  74. Krystal Maughan ORCID PhD student, University of Vermont
    Comment

    I have published in both Mathematics (Number theory, cryptography) and AI (not GenAI as used to "write proofs"; more so what I would call Machine Learning) and have seen "AI" as a useful tool for Mathematicians (e.g. work on Murmurations using LMFDB). However, I do agree that there should be oversight with respect to GenAI, having just attended a workshop at the intersection of AI and Number Theory. I'm writing as someone at the early stages of their career, seeing my peers struggle to find jobs after sending out hundreds of applications, and knowing some of the toxic and extractive parts of the AI world. One comment I heard years ago when Machine Learning first was applied to Mathematics was that "we don't want to feel steamrolled; it should be a collaboration". Becoming a competent mathematician takes time, dogged commitment and investment by others. It requires active work, humility and failing over and over again. I think a lot about what kind of legacy the senior mathematical researchers are leaving behind for us (the junior ones) to navigate, since present negotiations affect future yet-to-become mathematical researchers. I'm happy to see that we are having these discussions and am looking forward to future ones in this direction.

  75. Ricardo Menares ORCID Pontificia Universidad Católica de Chile
  76. Yohsuke Matsuzawa ORCID Associate professor, Osaka Metropolitan University
  77. Akhil Mathew ORCID University of Chicago
  78. Connor Olson verified email University of Washington
  79. Steven J. Hobson, MD FACC Mechanical Engineer, Clinical Cardiologist
  80. Atsushi Yoshikawa verified email Faculty of Mathematics, Kyushu University
  81. Claudio Gonzales ORCID Assistant Professor, Carleton College
  82. Scott Robson ORCID Research Data Analyst, Northwestern Medicine
    Comment

    We stand to lose much if we don’t think deeply about the impact of AI tools in the production of mathematical and scientific research and their intersection. Errors and verification costs are perhaps just the surface problem. Ethical questions are deeper concerns. Deeper still is we risk losing the ability to think intensely, even passionately, while teaching future generations that thinking is something that can be largely or completely outsourced.

  83. Agnishom Chattopadhyay ORCID Research Engineer, Imiron
  84. Peter Hintz ORCID Professor, Pennsylvania State University
  85. Sunay Joshi verified email University of Pennsylvania
  86. Deewang Bhamidipati ORCID Carleton College
  87. Mauricio Ayala-Rincon ORCID Full Professor, Universidade de Brasília
    Comment

    The Leiden Declaration outlines the principles for the fair and rigorous application of computational tools in mathematics, always crediting the work of our peers, providing certified and reproducible formalizations of our proofs, and, most importantly, preserving the culture of mathematical research.

  88. Kostiantyn Drach ORCID Associate professor, Universitat de Barcelona
  89. Joris Roos ORCID University of Massachusetts Lowell
  90. María Isabel Cortez ORCID Associate Professor, Pontificia Universidad Católica de Chile
  91. Leonardo Fernandes Guidi ORCID IME/UFRGS
  92. Kim Morrison ORCID Lean Focused Research Organization
    Comment

    I'm really excited about the future of mathematics, with incredible new capabilities and insights available from AI. The transformation is inevitable: we can do it badly, or we can do it well. Let's try to do it well! The Leiden Declaration provides the ethical and practical framework we need.

  93. Tim Hsu ORCID Professor, San Jose State Univ.
  94. Scott MacLachlan ORCID Professor, Memorial University of Newfoundland
  95. Daniel Quigley ORCID Indiana University Bloomington
  96. Lorenzo Maniscalco verified email University of Turin
  97. Daniel Disegni ORCID Mathematics, Aix-Marseille University
  98. Yu Shen verified email Michigan State University
  99. Sam Farnsworth verified email University of California, Los Angeles
  100. Jackson Walters ORCID Adjunct Faculty, Northern Virginia Community College