MAIN CONTENT
Seventeen million lines of Lean code and reactions from Fields medallists
On Friday, November 20, 2026 — as a conference opened in Shenzhen, southern China running through Monday for the 200th anniversary of German mathematician Bernhard Riemann's birth — Anthropic mathematician Levent Alpoge stated on his personal website that he solved the hypothesis using Claude after six weeks of code generation, releasing roughly 17 million lines of code in the open-source Lean language and interactive theorem prover rather than an unreadable manuscript while rejecting cynical timing speculation following Anthropic's Nasdaq listing.
The announcement sent shock waves through mathematicians attending and following the event:
Fields medallist Shing-Tung Yau — director of Tsinghua University's Yau Mathematical Sciences Centre and scientific committee chair for the bicentennial conference — told The Beijing News that mathematics has entered a new era shifting toward finding results first and examining them afterward like biology, physics, and chemistry, while noting human intelligence still cannot compare and problem-solving exceeds question-posing in machines.
Oxford professor and fellow Fields medallist James Maynard — a conference speaker who has contributed to Riemann hypothesis research — told The New York Times he believed the artificial intelligence proof was likely correct at around 98 or 99 percent certainty.
Fields medallist Terence Tao of the University of California, Los Angeles wrote in a blog post the same day that digesting the proof would take time, adding that having a human author to communicate with helps.
"It being obviously inappropriate to attempt to post some massive unreadable travesty of 'slop', I decided to only present Lean." Levent Alpoge
Significance of the Riemann hypothesis and path to formal recognition
First proposed in 1859 by Bernhard Riemann (1826–1866), born in September 1826, and unproven for 167 years, the Riemann hypothesis concerns prime-number distribution, anchors modern number theory, and ranks among the seven Millennium Prize Problems named by the Clay Mathematics Institute in 2000 carrying a one-million-dollar prize each — of which only one has been officially resolved so far.
Independent mathematicians began checking the Claude-generated Lean code — with early confirmations of mathematical accuracy reported — though translating the underlying logic into a human-readable paper and securing formal recognition will take time.
"This puts us in the rather extraordinary situation this age has made possible, in which I am presenting the world with a complete argument before having mastered a massive neighbourhood around its mathematics." Levent Alpoge















