The Riemann hypothesis has made fresh news. On November 20, 2026, Anthropic mathematician Levent Alpöge posted a statement on his personal website saying he had found a proof of the Riemann hypothesis together with Claude, and that the same method also covers the broader generalized Riemann hypothesis. The proof has been released as Lean code, but it has not yet been peer-reviewed — for now it can only be called a "claimed proof," not a proven one.
What exactly was delivered
Alpöge did not attach a human-readable paper, only a four-page note and the Lean code. Lean is a proof-assistant language that checks each step of a mathematical argument mechanically, and this code was released in 14 archives. In the note, Alpöge admits he has not yet fully mastered all the mathematics involved, and the handwritten version is still to come; before publishing, he consulted Anthropic CEO Dario Amodei. MIT mathematician Andrew Sutherland has already compiled the Lean code independently, saying "the world has changed," while stressing that much work remains to understand the mathematical details. Compiling is not the same as confirming the claim: whether the formalized statement truly corresponds to the Riemann hypothesis still needs to be checked by experts, step by step. About a week earlier, Anthropic reported that an unreleased Claude research version had raised the lower bound on the fraction of zeta-function zeros on the critical line from 41.6% to 67.2%, validated by two in-house mathematicians. The timing is also striking: the statement came the day after Anthropic's market debut.
Why a 167-year-old conjecture matters
The Riemann hypothesis was proposed by German mathematician Bernhard Riemann in 1859. It asks whether there is an exact pattern behind the distribution of prime numbers: whether all non-trivial zeros of the Riemann zeta function lie on a single critical line. It is one of the seven Millennium Prize Problems, carrying a $1 million bounty from the Clay Mathematics Institute, and in 167 years no one has proved or disproved it. In October, a batch of mathematical results released by OpenAI made progress on prime numbers, and AI's mathematical abilities keep setting new records — Gemini 4 Carbon has just topped a math leaderboard. If the proof holds, many results that assume the Riemann hypothesis would become theorems at once, and branches of number theory that were only partly connected would gain a shared "dictionary" — one mathematician called it a "Rosetta Stone." The limits matter too: this has no direct real-world application. It will not change bank encryption or make the internet faster.
Verification will not take long. With the Lean code public, number theorists and the Lean community can check the compilation and the formal statements within days; the next key moment will be statements from the Clay Mathematics Institute and leading authorities, and the timing of the human-readable version Alpöge has promised will decide how quickly the field can absorb the method. Until then, the accurate wording remains "claimed proof," not "the Riemann hypothesis has been solved.