Anthropic Mathematician Claims AI Breakthrough on 167-Year-Old Riemann Hypothesis

Graciela Maria Reporter

| 2026-11-21 09:18:25

An Anthropic researcher has sparked intense debate across the global mathematics and technology sectors after claiming to have solved the Riemann Hypothesis—a foundational math problem that has baffled scholars for 167 years—with the help of the company’s AI model, Claude.

Levent Alpoge, a mathematician working at Anthropic, published a brief four-page summary on his personal website on Friday detailing the purported breakthrough. Alongside the summary, Alpoge released formal proof files coded in Lean, a specialized programming language designed for automated theorem verification. He asserts that the collaborative effort resolved both the original 1859 conjecture by Bernhard Riemann and the broader Generalized Riemann Hypothesis.

Despite the bold claim, the mathematical world is urging caution until rigorous peer review can be conducted. Brian Conrey, a prominent specialist in the field, noted that while a valid proof would represent a watershed moment in mathematical history, he has only reviewed Alpoge's online summary rather than the complete work. Similarly, University of Bath professor Kit Yates emphasized that the proof will face exhaustive scrutiny, noting that confirmation would fundamentally reshape perceptions of artificial intelligence's analytical power.

Beyond the mathematical implications, the methodology behind the alleged proof has drawn significant attention. Alpoge revealed that his collaboration with Claude generated over 250,000 pages of material, admitting that the AI eventually outpaced his own comprehension as it finalized the argument. New York University visiting professor Nestor Guillen viewed this transparency favorably, suggesting it highlights AI's potential as an advanced research assistant rather than a replacement for human intellect.

The announcement arrives amid escalating competition among leading AI labs to conquer high-profile scientific challenges, closely following OpenAI’s recent claims regarding the Navier-Stokes equations—a dispute that previously sparked public friction between Alpoge and OpenAI CEO Sam Altman over data usage. Furthermore, Alpoge acknowledged potential skepticism regarding the timing of his release, which occurred just 24 hours after Anthropic’s landmark debut on the U.S. stock market.

Central to understanding the distribution of prime numbers, the Riemann Hypothesis posits that all non-trivial zeros of the Riemann zeta function share a real part of 1/2. Designated as one of the seven Millennium Prize Problems by the Clay Mathematics Institute in 2000, the conjecture carries a $1 million reward. Under the institute's strict guidelines, Alpoge's proof must be formally published and withstand at least two years of independent verification and consensus from the broader mathematical community before any prize is awarded.

[ⓒ Global Economic Times. 무단전재-재배포 금지]

WEEKLY HOT