OpenAI announced in May 2026 that its new general-purpose reasoning model has produced an original mathematical proof disproving a geometry conjecture first posed by Paul Erdős in 1946 — and this time, independent mathematicians are vouching for it.
The company says its model disproved a belief that had stood for nearly eight decades: that the best possible solutions to the problem resembled square grids. According to OpenAI, the model discovered “an entirely new family of constructions that performs better.” OpenAI called it “the first time AI has autonomously solved a prominent open problem central to a field of mathematics.”
The announcement carries extra weight given OpenAI’s recent history on this topic. Seven months earlier, the company’s then-VP Kevin Weil posted on X that GPT-5 had solved ten previously unsolved Erdős problems. That claim unraveled when it emerged the model had only surfaced solutions already existing in published literature. Weil deleted the post after criticism from figures including Meta’s Yann LeCun and Google DeepMind CEO Demis Hassabis.
This time, OpenAI published supporting statements from several mathematicians, including Noga Alon, Melanie Wood, and Thomas Bloom, who runs the Erdős Problems website. Bloom had previously described Weil’s original post as “a dramatic misrepresentation.” His endorsement of the new result is notable given that context.
OpenAI emphasized that the proof came from a general-purpose reasoning model — not a system built specifically for mathematics — suggesting AI may now be capable of sustaining complex chains of reasoning and connecting ideas across disciplines. The company said this could have implications for biology, physics, engineering, and medicine.
“AI is helping us to more fully explore the cathedral of mathematics we have built over the centuries,” Bloom said in a statement. “What other unseen wonders are waiting in the wings?”
Source: TechCrunch