Riemann Hypothesis: Two mathematicians claim proof of 167-year-old problem

Share:
Audio Loading voice…
Riemann Hypothesis: Two mathematicians claim proof of 167-year-old problem

Synopsis

For the first time, two competing proofs of the 167-year-old Riemann hypothesis have emerged on the same day — one built with Anthropic's Claude AI generating 17.4 million lines of formal Lean code, the other a human-reasoned approach from Stanford now under scrutiny. Neither is confirmed, but the race itself marks a historic moment for mathematics and artificial intelligence.

Key Takeaways

Levent Alpöge of Anthropic and Alessandro Maria Masullo of Stanford University both claimed proofs of the Riemann hypothesis on the same day.
Alpöge used Anthropic's Claude AI and released a 14-part archive with nearly 17.4 million lines of Lean code for formal verification.
Masullo's proof covers the Generalised Riemann Hypothesis (GRH) and was developed without AI, but has attracted queries from reviewers — some of whom used AI tools to identify issues.
The Riemann hypothesis was first proposed by Bernhard Riemann in 1859 and concerns the distribution of prime numbers via the Riemann zeta function .
Proving it is one of the Clay Mathematics Institute's seven Millennium Prize Problems , each worth $1 million , requiring at least two years of peer scrutiny before any prize can be awarded.

Two mathematicians have independently claimed to have solved the Riemann hypothesis — a 167-year-old unsolved problem considered one of the most consequential in mathematics — setting up an extraordinary parallel contest between an AI-assisted proof and one developed through classical human mathematical reasoning.

The Two Competing Claims

Levent Alpöge, a mathematician at AI company Anthropic, announced earlier this week that he had arrived at a proof of the Riemann hypothesis with assistance from Anthropic's Claude AI model. On the same day, Alessandro Maria Masullo, a researcher at Stanford University, submitted a separate proof of the Generalised Riemann Hypothesis (GRH) — a broader formulation of the original conjecture. Stanford mathematician Ravi Vakil described Masullo's work as having been developed without AI assistance.

Where Each Proof Stands

Alpöge's approach has reportedly cleared several important computer-based verification checks. He released a four-page note alongside a 14-part archive containing nearly 17.4 million lines of Lean code — a formal mathematical language used to mechanically verify the logical steps of a proof. His work also reportedly establishes the Generalised Riemann Hypothesis for Dirichlet L-functions.

Masullo's proof, by contrast, has attracted scrutiny from mathematicians who have identified queries in the draft. Notably, the problems flagged in Masullo's work were themselves identified with the assistance of different AI models — a detail that underscores how deeply AI tools have permeated even the process of peer evaluation. Vakil acknowledged that the issues would need to be resolved before Masullo's proof could be accepted by the mathematical community.

What the Riemann Hypothesis Actually Says

The hypothesis was proposed by German mathematician Bernhard Riemann in 1859 as a theory about the distribution of prime numbers — integers such as 2, 3, 5, 7, and 11 that are divisible only by themselves and 1. Riemann's conjecture attempts to describe the precise pattern in which these numbers appear, via a mathematical construct known as the Riemann zeta function. Despite over a century and a half of attempts by the world's leading mathematicians, no proof has ever been universally accepted.

The Clay Prize and the Road to Verification

Proving the Riemann hypothesis is one of the seven Millennium Prize Problems identified by the Clay Mathematics Institute in 2000, each carrying a prize of $1 million. However, the bar for claiming the prize is deliberately high: a proposed proof must undergo at least two years of scrutiny after publication in a qualifying peer-reviewed journal, and must achieve broad acceptance among the global mathematics community before any award is made.

This is not the first time claims of a Riemann proof have emerged — several prior attempts over the decades were eventually found to contain errors. Whether either of these two new submissions survives the rigorous verification process remains to be seen, and mathematicians worldwide are expected to subject both to intensive examination in the weeks ahead.

Point of View

One human-reasoned — is less a resolution of the Riemann hypothesis than a stress test for how mathematics itself validates knowledge in the AI era. That reviewers are using AI models to find flaws in an AI-assisted proof creates a recursive dynamic the Clay Institute's two-year scrutiny rule was never designed to handle. The deeper question is not whether either proof is correct today, but whether the field's verification infrastructure — peer review, journal timelines, community consensus — is fast or robust enough to keep pace with AI-accelerated mathematical output. If Alpöge's Lean-formalised approach holds, it would also mark the first time a Millennium Prize problem was cracked with substantial AI involvement, resetting expectations for every unsolved problem that follows.
NationPress
21 Nov 2026

Frequently Asked Questions

What is the Riemann hypothesis?
The Riemann hypothesis is a conjecture proposed by German mathematician Bernhard Riemann in 1859 about the distribution of prime numbers, described through a mathematical construct called the Riemann zeta function. It has remained unproven for 167 years and is considered one of the most important unsolved problems in mathematics.
Who has claimed to prove the Riemann hypothesis?
Two mathematicians made competing claims on the same day: Levent Alpöge of Anthropic, using Claude AI assistance, and Alessandro Maria Masullo of Stanford University, working without AI. Both claims are currently under review by the broader mathematics community.
What is Lean code and why does it matter here?
Lean is a formal mathematical language that allows proofs to be expressed in a way that computers can mechanically verify step by step. Alpöge released nearly 17.4 million lines of Lean code alongside his claim, which has reportedly cleared several computer-based verification checks — a significant, though not final, signal of validity.
Has the Riemann hypothesis been 'solved' before?
Several mathematicians over the decades have announced proofs that were later found to contain errors. Neither of the current claims has yet been accepted; Masullo's proof has already attracted queries, and both must survive at least two years of peer scrutiny in a qualifying journal before any Clay Millennium Prize could be awarded.
What prize is at stake for proving the Riemann hypothesis?
The Clay Mathematics Institute designated the Riemann hypothesis as one of seven Millennium Prize Problems in 2000, each carrying a $1 million prize. A proof must be published in a peer-reviewed journal, withstand at least two years of scrutiny, and gain broad acceptance among mathematicians before the prize is awarded.
Nation Press
The Trail

Connected Dots

Tracing the thread behind this story — newest first.

8 Dots
  1. Latest 4 months ago
  2. 6 months ago
  3. 6 months ago
  4. 6 months ago
  5. 7 months ago
  6. 1 year ago
  7. 1 year ago
  8. 1 year ago
Google Prefer NP
On Google