Decades-Old Mathematical Riddle Finally Solved, Shaking the Mathematics World
Jakarta, CNBC Indonesia – The global mathematics community is in turmoil after OpenAI simultaneously released more than 350 findings generated by artificial intelligence (AI) on Tuesday (6 October 2026). The release has prompted mixed reactions among mathematicians, ranging from admiration to concerns about the future of the discipline.
According to The New York Times, the tens of thousands of pages of findings published by OpenAI include answers to significant open questions across almost every branch of mathematics, including algebra, number theory, theoretical computer science, mathematical logic and topology.
The finding attracting the most attention is a proof of the “quasi-Riemann hypothesis”, a conjecture related to the Riemann Hypothesis, one of the best-known problems on the list of Millennium Prize Problems. Alex Kontorovich, head of the mathematics department at Rutgers University, said the achievement would immediately earn a Fields Medal, the highest honour in mathematics, if it had been produced by a human.
Even so, none of the findings fully solves any of the five remaining Millennium Prize Problems. The collection is, however, said to contain related advances on all five.
Martin Bridson, president of the Clay Mathematics Institute and an Oxford mathematician, described OpenAI’s release as astonishing. Until recently, he said, it would have been impossible to imagine the boundaries of mathematical knowledge shifting so far in a single day.
Alongside the admiration, some academics have raised concerns about verification. Bryna Kra, a former president of the American Mathematical Society, called the results extraordinary but said the tens of thousands of pages now needed to be examined carefully. She noted that the process had bypassed the peer-review mechanism that has traditionally served as the standard for verifying scholarly work.
Kra also fears that releases of this kind could damage the research ecosystem that made such work possible. She said mathematicians using AI tools risk losing control of their own ideas.
Concerns about accuracy are not unfounded. By Thursday (8 October 2026), OpenAI had updated its online repository, withdrawing three papers and correcting several others. An OpenAI spokesperson said the company would promptly correct errors that were identified and withdraw papers if no viable correction could be made, adding that the process was iterative, like other research manuscripts.
Ken Ono, a professor at the University of Virginia and a founding mathematician at Axiom Math, said many of his colleagues were initially shaken by the release. He added that the feeling quickly gave way to enthusiasm and a determination to aim higher.
Meanwhile, the legendary French mathematician Jean-Pierre Serre, who turned 100 last month, discovered that one of OpenAI’s findings proved a conjecture he had proposed nearly 70 years ago. The Fields Medal and Abel Prize winner said he was in a dilemma. AI, he said, enriches the pleasure of learning but diminishes the satisfaction of making a new discovery—a consequence he considers most serious for young mathematicians.
Doctoral students feel the greatest impact
The upheaval is thought to be felt most acutely by postgraduate students. Thomas Carlson, a doctoral student at Montana State University, questioned how the value of a mathematics doctorate awarded in 2028 could be kept on a par with those awarded several years earlier.
Yuanning Zhang, a doctoral student at Northwestern University, said he felt more dizzy than judgemental about the developments. If the findings are verified, he said, AI-assisted mathematical research appears to be starting to operate on an industrial scale.
Amid the debate, reports have circulated that OpenAI will soon release another wave of findings.