World’s top 25 Fields Medalists warn machine proofs are sabotaging hardest math

A group of the world’s most decorated mathematicians is warning that the race to make AI solve difficult mathematics could damage the field it aims to advance. Twenty-five Fields Medalists have signed an open letter criticizing the growing push by AI companies to solve famous mathematical problems as demonstrations of model capability. The signatories argue that a correct answer alone does not capture what makes mathematical research valuable. Their concern goes beyond competition over who solves a problem first. They say rushed agentic mathematics proofs could disrupt how mathematicians verify discoveries, develop new ideas, and pass knowledge to future generations. The warning arrives as AI systems make increasingly strong claims about solving difficult mathematical problems. OpenAI’s recent proof, for instance, has yet to receive verification from the wider mathematical community. More than answers A mathematical proof can settle a question, but researchers also care about the machinery behind it. New techniques can reveal relationships between seemingly unrelated areas and create tools that other mathematicians later build upon. That process can take years. Researchers discuss results, challenge assumptions, simplify arguments, and connect discoveries with earlier work. Eventually, an initially obscure idea can become part of standard mathematical education. The letter’s authors fear that AI companies could short-circuit that process. Systems can produce an apparent solution quickly, but mathematicians still need time to establish whether it works and determine what the underlying method actually teaches. Attribution presents another problem. Agentic mathematics solutions may draw upon decades of existing research, making it harder to determine where genuinely new ideas originated. “Often these solutions are announced in a rush,” the mathematicians wrote, citing the lack of time for careful write-ups and proper attribution. They warned that AI-generated ideas need mathematicians to develop and integrate them into the broader body of knowledge. Race could change research The economic scale of AI development adds another layer to the dispute. Frontier labs can devote enormous computing resources toward mathematical challenges that once required years of human effort. That creates an unusual incentive. If researchers believe an unpublished idea could help an AI system produce a major result, they may become less willing to share their work openly. Some mathematicians have already questioned whether using coding tools from AI companies could expose their research to future model development. Such concerns could weaken the open exchange that has traditionally helped mathematics progress. The dispute has also spilled beyond academic discussions. OpenAI withdrew its sponsorship of a mathematics event at Caltech on Thursday after researchers criticized the company. A separate initiative, the Leiden Declaration, has also called for changes in how mathematicians, institutions and policymakers approach increasingly capable mathematical AI. The signatories are not rejecting the technology outright. They see clear potential for AI to accelerate mathematical research and help researchers explore difficult problems. Their warning concerns what happens when solving the problem becomes the objective itself. Mathematics could gain faster answers while losing some of the human process that turns those answers into lasting knowledge. The question now extends beyond mathematics. As AI systems take on more intellectual work, other professions may face the same tension between producing an answer and understanding why that answer matters. Get the latest in engineering, tech, space & science - delivered daily to your inbox.Aamir is a seasoned tech journalist with experience at Exhibit Magazine, Republic World, and PR Newswire. With a deep love for all things tech and science, he has spent years decoding the latest innovations and exploring how they shape industries, lifestyles, and the future of humanity.