Twenty-five distinguished mathematicians, all recipients of the Fields Medal, have publicly expressed concerns about the growing role of AI labs in mathematical research. They argue that the competitive drive among AI developers to solve famous math problems risks undermining the traditional values of attribution and collaborative scholarship.
Recently, NYU professor Tristan Buckmaster accused OpenAI of pressuring him to omit credit for a collaborator affiliated with Anthropic, who contributed to solving a significant mathematical problem. Buckmaster also questioned whether OpenAI leveraged insights from their Codex model to independently generate a major proof during an intensive inference session.
In response to criticism from CalTech researchers, OpenAI withdrew its sponsorship of a mathematics event at the university. This incident underscores the tension between AI developers and the academic community.
The mathematicians emphasize that while AI-generated solutions have the potential to advance human knowledge, these breakthroughs must be clearly communicated, verified, and integrated into the broader mathematical framework. They caution against rushed announcements that lack proper documentation and fail to acknowledge prior work, which raises concerns about plagiarism and attribution.
There is growing unease within the community that AI labs might be incorporating researchers' work—such as code and proofs developed using Codex—into their own models without consent. This dynamic could incentivize secrecy and hinder the open research culture that has traditionally driven mathematical progress.
This open letter builds on the Leiden Declaration, a statement released earlier this year by mathematicians addressing the implications of AI-generated proofs and proposing guidelines for researchers and institutions.
The mathematicians argue that the true value of their work lies not only in the proofs themselves but also in the intellectual ecosystem that nurtures new ideas, educates students, and connects discoveries to broader human knowledge.
They warn that the challenges faced by the mathematical community reflect broader issues confronting many scientific and creative fields as AI reshapes workflows. Ensuring that the purpose and integrity of scholarly work are preserved amid these changes is a concern for all disciplines.