Twenty-five prominent mathematicians signed an open letter arguing that AI labs are threatening their intellectual work as they seek to outdo each other with solutions to famous math problems. Each signer has received the Fields Medal, considered the most prestigious award in mathematics. This week, New York University professor Tristan Buckmaster accused OpenAI of pressuring him not to credit a contributor working for Anthropic for solving an important math problem, and questioned whether the company had used his work with Codex to produce its own groundbreaking proof during a marathon weekend of inference. On Thursday, OpenAI pulled its sponsorship of a math event at CalTech after researchers at the university criticized the company. While the ability of AI models to solve the world’s biggest mathematical challenges could be a boon to humanity, the signatories of the new letter argue that will only be the case if those solutions can be understood and communicated by the mathematics community and, ultimately, the rest of the world. “Often these solutions are announced hastily, leaving no time for proper writing, isolation of new methods and ideas, and citation of relevant previous work by others,” they wrote, and the OpenAI test remains unverified. “As in all creative professions, this raises serious questions of attribution and plagiarism. Furthermore, without the willing mathematicians who must take charge of their development and integration into the mathematical canon, the ideas conceived by AI would never come to full life and the crucial chain of human transmission between mathematicians would be lost.” While other mathematicians become paranoid and wonder if their use of Codex was in turn incorporated into the new OpenAI models, there is a real fear that the culture of open research is being threatened. Today, if frontier labs see a useful path to a discovery, they can spend tens of millions of dollars using LLM to beat the original researchers to a test, a dynamic that will incentivize secrecy. This letter follows the Leiden Declaration, published by a working group of mathematicians in June. That paper also addresses the ways in which LLM testing will change your work and offers a set of recommendations for mathematicians, institutions and policymakers. As with software engineering and other areas where AI tools are changing workflows, mathematicians find justification in the work that surrounds the work: the value in mathematics is not just the proofs and who gets the credit, but the intellectual superstructure that nurtures students, finds new questions and ideas, and integrates them into broader human civilization. And if you’re not particularly interested in the cutthroat world of high-stakes mathematical proofs, don’t forget: your field of interest is as follows. “The problems facing the mathematics community now are similar to those facing other scientific and creative professions, and indicate problems that all of humanity could face: how to ensure that, as AI changes the way work is done, we do not lose sight of what that work was intended to achieve in the first place,” they wrote. When you buy through links in our articles, we may earn a small commission. This does not affect our editorial independence.
