OpenAI withdraws Caltech Mathathon sponsorship after mathematician revolt

OpenAI has withdrawn its sponsorship of the Caltech "Mathathon," a research-level mathematics hackathon, following a coordinated revolt by the academic community. The decision comes after hundreds of mathematicians warned that the event's approach to AI-driven research could have "destructive impacts" on the field.

The Mathathon was designed as a high-stakes competition where teams were provided with $20,000 in AI tokens to solve open mathematical problems over a 40-hour window. OpenAI and Anthropic had initially pledged a combined $2 million in AI credits to the event, which also saw sponsorship from Cognition, Y Combinator, a16z, and DARPA's expMath program. However, the initiative sparked a backlash from roughly 500 signatories of an open letter, including Caltech mathematicians, who accused AI companies of promoting "slop mathematics"—results that appear plausible but remain shallow and lack rigorous proof or proper attribution.

The Clash Over Rigor

This conflict is rooted in a deepening divide between the "move fast" culture of AI labs and the exacting standards of the mathematical community. While some prominent figures have utilized AI for experimentation, many senior researchers argue that AI-generated proofs prioritize the appearance of problem-solving over genuine mathematical understanding. This sentiment was formalized in the Leiden Declaration on AI and Mathematics, a community statement warning against the commercialization of research and the tendency of AI firms to announce scientific results via press releases rather than through traditional, rigorous peer review.

Intellectual Property and Ethics

Beyond the debate over rigor, the dispute has touched on the ethics of data usage. Tristan Buckmaster, a mathematician at NYU, alleged that OpenAI may have used his private chatbot interactions to arrive at a solution for the Navier-Stokes equations. Such claims highlight a growing anxiety within the scientific community regarding the devaluation of human expertise and the potential for AI companies to claim breakthroughs based on the uncredited intellectual labor of researchers.

The Path Forward

In response to the backlash, OpenAI expressed a desire to engage more deeply with the community on how to integrate the technology and communicate its impacts. For now, the event serves as one of the first organized, credentialed pushbacks from a scientific discipline against the influence of major AI labs. Observers will be watching to see if this leads to a formal framework for AI-human collaboration in mathematics or if the rift between commercial labs and academia continues to widen.