{"id":63031,"date":"2026-11-21T09:57:56","date_gmt":"2026-11-21T06:57:56","guid":{"rendered":"https:\/\/infolight.ua\/?p=63031"},"modified":"2026-11-21T09:57:56","modified_gmt":"2026-11-21T06:57:56","slug":"anthropic-mathematician-claude-riemann-hypothesis-proof-claim","status":"publish","type":"post","link":"https:\/\/infolight.ua\/en\/2026\/11\/21\/anthropic-mathematician-claude-riemann-hypothesis-proof-claim\/","title":{"rendered":"Anthropic mathematician says he and Claude have proved the Riemann hypothesis: what we know and what is still being checked"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">On <strong>20 November<\/strong>, Anthropic mathematician <strong>Levent Alp\u00f6ge<\/strong> announced that, working with the company\u2019s AI model Claude, he has found a proof of the <strong>Riemann hypothesis<\/strong> \u2013 the most famous unsolved problem in mathematics, first posed in 1859. The proof has not been released as a paper but as a vast archive of code in Lean, a language in which a computer checks every step. It has passed its first independent checks, yet mathematicians stress that the problem will not be considered solved for some time.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Alp\u00f6ge posted a four-page note on his website, <a href=\"https:\/\/alpo.ge\/remark.pdf\">alpo.ge<\/a>, reports <strong><a href=\"https:\/\/www.scientificamerican.com\/article\/anthropic-mathematician-claims-to-prove-riemann-hypothesis-most-famous-problem-in-all-of-math\/\">Scientific American<\/a><\/strong>. He says the same method also proves the <strong>Generalized Riemann Hypothesis<\/strong> for Dirichlet L-functions \u2013 a broader statement on which hundreds of results in number theory depend.<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"wp-block-paragraph\">\u2018I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis,\u2019 the note begins.<\/p>\n<\/blockquote>\n\n\n\n<h2 class=\"wp-block-heading\">41 days and 239 megabytes of code<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">According to Alp\u00f6ge, the timeline of the work was compiled by Claude itself after searching their joint material. It covers <strong>41 days<\/strong>, from 6 October to 16 November 2026. Alp\u00f6ge says he set the direction and brought together tools from different areas of number theory, while Claude did most of the heavy technical derivation and, by the final stage, was moving faster than he was, <a href=\"https:\/\/www.zmescience.com\/science\/math\/riemann-hypothesis-solved-claude\/\">ZME Science<\/a> reports.<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"wp-block-paragraph\">\u2018Claude flew, and danced, and sparkled near that finish line,\u2019 Alp\u00f6ge writes of the final weeks.<\/p>\n<\/blockquote>\n\n\n\n<p class=\"wp-block-paragraph\">The proof comes as a 14-part archive containing <strong>175,187 files<\/strong> and roughly 17.4 million lines of Lean \u2013 <strong>239 MB<\/strong> in total. A machine-generated prose version reportedly runs to more than 250,000 pages, so Alp\u00f6ge chose to release only the Lean code rather than a text no human could read, and has promised to simplify and explain the argument himself.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">For Alp\u00f6ge, it is also a personal moment: he says he wanted to tackle the hypothesis since his student days and now loses \u2018a principal guiding light\u2019 of his research. At the same time, he calls what he has seen the dawn of \u2018something far greater\u2019 across science.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">What the Riemann hypothesis says<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">In November 1859 the German mathematician <strong>Bernhard Riemann<\/strong> sent the Prussian Academy of Sciences a paper titled \u2018On the Number of Primes Less Than a Given Magnitude\u2019. Prime numbers \u2013 2, 3, 5, 7, 11, 13 and so on \u2013 can be divided only by 1 and themselves. They look randomly scattered, but Riemann showed that their distribution is encoded in a single function: the zeta function \u03b6(s). A fuller introduction is available on <a href=\"https:\/\/en.wikipedia.org\/wiki\/Riemann_hypothesis\">Wikipedia<\/a>.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><a href=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/riemann-1859-zvit-akademii.jpg\"><img loading=\"lazy\" decoding=\"async\" width=\"1060\" height=\"990\" src=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/riemann-1859-zvit-akademii.jpg\" alt=\"The opening of the Prussian Academy of Sciences report for November 1859, announcing Riemann\u2019s paper on the number of primes below a given magnitude, in German\" class=\"wp-image-63032\" srcset=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/riemann-1859-zvit-akademii.jpg 1060w, https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/riemann-1859-zvit-akademii-1000x934.jpg 1000w, https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/riemann-1859-zvit-akademii-768x717.jpg 768w\" sizes=\"auto, (max-width: 1060px) 100vw, 1060px\" \/><\/a><figcaption class=\"wp-element-caption\">The Prussian Academy\u2019s report for November 1859, announcing Riemann\u2019s paper on the number of primes. Image: <a href=\"https:\/\/commons.wikimedia.org\/wiki\/File:RiemannPrim1859.djvu\">Wikimedia Commons<\/a>, public domain<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">For real s greater than 1, the zeta function is the simple sum \u03b6(s) = 1 + 1\/2<sup>s<\/sup> + 1\/3<sup>s<\/sup> + 1\/4<sup>s<\/sup> + \u2026, and Leonhard Euler had already noticed that the same value is given by a product over all primes. Riemann extended the function to the whole complex plane \u2013 numbers of the form s = a + bi \u2013 and focused on the points where it equals zero.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">There are two kinds of zeros. The \u2018trivial\u2019 ones sit predictably at the negative even numbers: \u22122, \u22124, \u22126 and so on. The rest, the \u2018non-trivial\u2019 zeros, of which there are infinitely many, lie in a narrow vertical strip between the lines Re(s) = 0 and Re(s) = 1. <strong>The Riemann hypothesis states that all of them lie exactly in the middle \u2013 on the \u2018critical line\u2019 Re(s) = 1\/2.<\/strong><\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><a href=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-rimana-kompleksna-ploshchyna.jpg\"><img loading=\"lazy\" decoding=\"async\" width=\"821\" height=\"820\" src=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-rimana-kompleksna-ploshchyna.jpg\" alt=\"Colour map of the Riemann zeta function on the complex plane, with dark points marking zeros lined up vertically along the line with real part 1\/2\" class=\"wp-image-63026\" srcset=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-rimana-kompleksna-ploshchyna.jpg 821w, https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-rimana-kompleksna-ploshchyna-768x767.jpg 768w\" sizes=\"auto, (max-width: 821px) 100vw, 821px\" \/><\/a><figcaption class=\"wp-element-caption\">The zeta function on the complex plane, with colour showing the function\u2019s value. The dark points on the vertical line near zero are non-trivial zeros at heights 14.13, 21.02, 25.01\u2026 Image: Jan Homann \/ <a href=\"https:\/\/commons.wikimedia.org\/wiki\/File:Complex_zeta.jpg\">Wikimedia Commons<\/a>, public domain<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">These zeros govern the \u2018noise\u2019 in how the primes are distributed. If the hypothesis holds, the number of primes up to any x deviates from its average estimate by no more than roughly \u221ax\u00b7ln x, and gaps between consecutive primes cannot grow too large. In practice the gaps are very uneven: some are 2, some 4, and in 2017 Martin Raab found a <a href=\"https:\/\/primegap-list-project.github.io\/record\/2017\/09\/04\/new-record-prime-gap-from-martin-raab\">gap of 6,582,144<\/a> between two probable primes with 216,841 digits each, with every number in between composite.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><a href=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-na-krytychnii-priamii.png\"><img loading=\"lazy\" decoding=\"async\" width=\"1280\" height=\"596\" src=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-na-krytychnii-priamii.png\" alt=\"Plot of the real part (red) and imaginary part (blue) of the zeta function along the critical line; zeros occur where both curves cross zero at the same point\" class=\"wp-image-63028\" srcset=\"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-na-krytychnii-priamii.png 1280w, https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-na-krytychnii-priamii-1000x466.png 1000w, https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/dzeta-funktsiia-na-krytychnii-priamii-768x358.png 768w\" sizes=\"auto, (max-width: 1280px) 100vw, 1280px\" \/><\/a><figcaption class=\"wp-element-caption\">Real (red) and imaginary (blue) parts of \u03b6(1\/2 + ix). Non-trivial zeros sit where both curves cross zero at once. Plot: Slonzor \/ <a href=\"https:\/\/commons.wikimedia.org\/wiki\/File:RiemannCriticalLine.svg\">Wikimedia Commons<\/a>, public domain<\/figcaption><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">The hypothesis cannot be settled by brute force. Computers have confirmed that more than 12 trillion of the first non-trivial zeros lie on the critical line (<a href=\"https:\/\/arxiv.org\/abs\/2004.09765\">a rigorous check by Dave Platt and Tim Trudgian<\/a> up to height 3\u00b710<sup>12<\/sup>), but there are infinitely many zeros, and a single exception would disprove it. That is why David Hilbert put it on his 1900 list of the century\u2019s great problems, and why in 2000 the Clay Mathematics Institute made it one of seven Millennium Prize Problems, each worth <strong>$1 million<\/strong>.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The hypothesis also has an unexpected relative in physics. In the early 1970s mathematician Hugh Montgomery and physicist Freeman Dyson noticed that the spacing between zeta zeros follows the same statistics as energy levels in heavy atomic nuclei and in models of quantum chaos. Why pure abstraction mirrors the physical world so closely is still unexplained.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Who is checking the proof, and how<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Lean checks each logical step mechanically, but a successful build answers only part of the question. Experts need to confirm that the code does not rely on hidden assumptions (extra axioms) and that the formal statement really is the Riemann hypothesis rather than something that merely looks like it, <a href=\"https:\/\/lean4.dev\/proofs\/riemann-hypothesis-claim\">lean4.dev<\/a> explains.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The first results came the same day. MIT mathematician <strong>Andrew Sutherland<\/strong> compiled the archive independently: according to <a href=\"https:\/\/kingy.ai\/blog\/levent-alpoge-claude-riemann-hypothesis-proof\/\">Kingy AI<\/a>, all 175,187 modules passed Lean\u2019s kernel check, and the headline theorems rely only on the standard axioms. Sutherland called it a \u2018historic moment\u2019 while stressing that much work remains on the mathematical details. Mathematician Elliot Glazer <a href=\"https:\/\/x.com\/ElliotGlazer\/status\/2123792743503933774\">wrote<\/a> that the formalised statement is indeed the Generalized Riemann Hypothesis.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The irony is that Sutherland is the author of a line, given to <a href=\"https:\/\/www.scientificamerican.com\/article\/the-riemann-hypothesis-is-a-million-dollar-math-problem-hardly-anyone-is-trying-to-solve\/\">Scientific American<\/a> back in 2024, that everyone is now quoting:<\/p>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p class=\"wp-block-paragraph\">\u2018An AI that can prove the Riemann hypothesis is not one I\u2019d want to meet. If that happens, mathematicians having jobs will be the least of our problems.\u2019<\/p>\n<cite>Andrew Sutherland, MIT<\/cite><\/blockquote>\n\n\n\n<p class=\"wp-block-paragraph\">Others are more guarded. \u2018Let\u2019s see if it compiles first,\u2019 said <strong>Kevin Buzzard<\/strong> of Imperial College London. Fields medallist <strong>Terence Tao<\/strong> wrote on his <a href=\"https:\/\/terrytao.wordpress.com\/\">blog<\/a> that the Lean community\u2019s <a href=\"https:\/\/leanprover.zulipchat.com\/#narrow\/channel\/113488-general\/topic\/Discussion.3A.20Alp.C3.B6ge.20Proves.20the.20Riemann.20Hypothesis\">Zulip thread<\/a> is probably the best place to follow the verification effort. \u2018It is both a wonderful time and a terrible time to be a research mathematician,\u2019 said James McKee of Royal Holloway, University of London.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Even if everything checks out, the prize will not come quickly: Clay\u2019s rules require publication, at least two years of scrutiny and general acceptance by the community. For now, the hypothesis officially remains open.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">The year AI took on the Millennium Problems<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">Alp\u00f6ge\u2019s claim caps a year of striking AI results in mathematics. In August, Anthropic <a href=\"https:\/\/www.anthropic.com\/research\/riemann-zeta\">reported<\/a> that a research version of Claude, asked to \u2018take a real stab\u2019 at the Riemann hypothesis, had not proved it but had raised the proven share of zeros on the critical line from <strong>41.6% to 67.2%<\/strong>. That result was checked by Alp\u00f6ge and Ralph Furman.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">On 8 September OpenAI <a href=\"https:\/\/openai.com\/index\/navier-stokes-solution\/\">announced<\/a> a solution to another Millennium Problem, on the Navier\u2013Stokes equations of fluid flow, produced by roughly 10,000 AI agents working in parallel over 88 hours. The company says it started after hearing rumours later linked to Alp\u00f6ge and mathematician Tristan Buckmaster, and credit for that work is still disputed. In October OpenAI released hundreds of new results, including a <a href=\"https:\/\/quantamagazine.org\/ai-proved-the-quasi-riemann-hypothesis-what-happens-now-20261019\">proof of the \u2018quasi-Riemann hypothesis\u2019<\/a>, a weaker version of the problem.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Alp\u00f6ge describes a different model: the human set the strategy, and the AI sped up and finished the technical work. Whether that becomes a template for science will depend on whether people can understand the proof, not just confirm that it is correct.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Does this threaten encryption?<\/h2>\n\n\n\n<p class=\"wp-block-paragraph\">On social media the news is already being tied to the end of cryptography: RSA\u2019s security rests on how hard it is to factor large numbers into primes, and the Riemann hypothesis is about primes too. But there is no direct link. A proof of the hypothesis gives no fast way to find factors \u2013 it confirms what mathematicians already believed and have long assumed in many calculations. Mathematician Alon Amit <a href=\"https:\/\/parabola.unsw.edu.au\/sites\/default\/files\/2024-02\/vol57_no1_2.pdf\">dismissed<\/a> the popular claim that such a proof would endanger internet security as \u2018sheer nonsense\u2019 back in 2021.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Cryptocurrencies are not at risk either: Bitcoin signatures use an elliptic curve over a prime field, and their strength does not depend on the Riemann hypothesis. The real challenge to RSA and today\u2019s digital signatures is quantum computing, which is why the world is already moving to post-quantum algorithms.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">In the coming weeks the Lean community will continue checking the code, while mathematicians try to grasp the idea behind it. Alp\u00f6ge describes his method as a descendant of Stepanov\u2019s method, with the required multiplicity supplied by a chain of theta lifts between unitary groups. If the proof survives scrutiny, the most famous problem in mathematics will have been solved for the first time by a human working in tandem with artificial intelligence.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><em>Main image: portrait of Bernhard Riemann, 1863, <a href=\"https:\/\/commons.wikimedia.org\/wiki\/File:Georg_Friedrich_Bernhard_Riemann.jpeg\">Wikimedia Commons<\/a>, public domain. Illustrations: <a href=\"https:\/\/commons.wikimedia.org\/wiki\/File:RiemannPrim1859.djvu\">Riemann\u2019s 1859 paper<\/a>, Jan Homann, Slonzor \u2013 Wikimedia Commons, public domain.<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Levent Alp\u00f6ge has released 239 MB of Lean code that he says proves the 167-year-old conjecture about prime numbers. The first machine checks have passed, but mathematicians say acceptance is still a long way off.<\/p>\n","protected":false},"author":25,"featured_media":63022,"comment_status":"open","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"wds_primary_category":1690,"footnotes":""},"categories":[1690,1307],"tags":[],"class_list":["post-63031","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-neus","category-technology"],"jetpack_featured_media_url":"https:\/\/infolight.ua\/wp-content\/uploads\/2026\/11\/bernhard-riemann-portret-1863.jpg","_links":{"self":[{"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/posts\/63031","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/users\/25"}],"replies":[{"embeddable":true,"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/comments?post=63031"}],"version-history":[{"count":1,"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/posts\/63031\/revisions"}],"predecessor-version":[{"id":63035,"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/posts\/63031\/revisions\/63035"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/media\/63022"}],"wp:attachment":[{"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/media?parent=63031"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/categories?post=63031"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infolight.ua\/en\/wp-json\/wp\/v2\/tags?post=63031"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}