Science
TechCrunch

AI claimed a solution to the 167-year-old Riemann Hypothesis — and a human did, too

Source Entity

Tim Fernholz

November 22, 2026
AI claimed a solution to the 167-year-old Riemann Hypothesis — and a human did, too

Anthropic mathematician Levent Alpöge claims to have solved the 167-year-old Riemann Hypothesis using the AI model Claude. The proof, which remains unverified by the mathematical community, has been released as Lean code.

The Riemann Hypothesis: A New Frontier for AI

Levent Alpöge, a mathematician affiliated with the AI lab Anthropic, has stunned the global research community by claiming to have successfully proved the Riemann Hypothesis. This 167-year-old mathematical riddle, which posits a hidden structure underlying the distribution of prime numbers, is one of the seven Millennium Prize problems. Historically, the hypothesis has served as the ultimate test of mathematical rigor, with its resolution promising to unlock profound secrets about the fundamental nature of numbers.

The Role of Claude in Mathematical Proof

Unlike traditional mathematical breakthroughs that rely solely on human intuition and manual derivation, Alpöge asserts that his discovery was facilitated by Anthropic’s AI model, Claude. By leveraging machine intelligence to navigate the complexities of the proof, Alpöge has ignited a debate regarding the future of computational mathematics. The use of Lean—a formal proof assistant—to compile the solution suggests a shift toward machine-verified mathematics, where AI is not merely a tool for calculation but a collaborative partner in formal logic.

Timing and Corporate Context

The announcement arrived at a delicate intersection of corporate and academic milestones. Alpöge noted that he contacted Anthropic CEO Dario Amodei late Sunday evening regarding the findings, strategically delaying the public release until after the company’s recent IPO. This timing highlights the tension between the high-stakes world of AI commercialization and the slow, deliberate process of scientific peer review, placing Anthropic at the center of both technological and mathematical history.

The Burden of Verification

Despite the excitement surrounding the claim, the mathematical community remains cautious. Alpöge’s proof has not yet been independently verified, and Anthropic itself has maintained a stance of non-involvement, declining to comment on the announcement. In the realm of mathematics, a claim—no matter how revolutionary—is essentially a placeholder until it withstands the rigorous scrutiny of experts who must validate the Lean code and the underlying logic.

Broader Implications for Mathematics

If the proof is verified, it would represent a seismic shift in the field, potentially eclipsing previous, less refined AI efforts in mathematical reasoning. The Riemann Hypothesis is deeply connected to the security of modern encryption and the foundational structure of number theory. A verified proof would not only validate the utility of AI in solving the world’s most difficult puzzles but would also likely force a re-evaluation of how mathematicians approach conjectures that have eluded humanity for nearly two centuries.

Future Trends in AI-Augmented Research

Looking ahead, this event signals a potential turning point where AI-human collaboration becomes the standard for tackling 'unsolvable' problems. As models like Claude become more proficient at formal languages like Lean, we may see an acceleration in the discovery of proofs that were previously beyond the reach of human cognitive constraints alone. Whether this specific claim stands or falls, it undeniably marks the beginning of a new era in computational science.

Multiple Citing Sources

Verification Required?

Read the full report from the primary source

Go to TechCrunch