What every source agrees on
Levent Alpöge, a mathematician at Anthropic, wrote on Friday that he had, working with Claude, discovered a proof of the Riemann hypothesis, in a four-page note Scientific American20 November 2026Open the article See it in the source listNature21 November 2026Open the article See it in the source list. The hypothesis, first proposed in 1859, concerns the distribution of the primes and carries a $1 million Clay prize New Scientist20 November 2026Open the article See it in the source listAl Jazeera English21 November 2026Open the article See it in the source list. A day later the communities behind the Lean proof language and its Mathlib library certified the result, writing that they have full confidence in the statement and the proof Scientific American21 November 2026Open the article See it in the source list. Andrew Sutherland at MIT compiled the code without issues Scientific American21 November 2026Open the article See it in the source listMint21 November 2026Open the article See it in the source list. An unpublished written version ran past 250,000 pages; what he released is Lean code New Scientist20 November 2026Open the article See it in the source list. Alpöge said the method also extends to the generalised Riemann hypothesis Al Jazeera English21 November 2026Open the article See it in the source listNDTV21 November 2026Open the article See it in the source list. The same day Ravi Vakil, the president of the American Mathematical Society and a mathematician at Stanford, posted a rival preprint written without machine assistance by Alessandro Maria Masullo of Stanford Scientific American20 November 2026Open the article See it in the source listNature21 November 2026Open the article See it in the source list.
What the sources contradict each other on
The numbers are the sources this piece is built from, listed in full at the foot of the page. Tap one to jump to it.
Whether the machine proof is settled is the first dispute.
The certifying communities announced full confidence, with independent researchers compiling the code without issues Scientific American21 November 2026Open the article See it in the source list, whereas Terence Tao said verification remains some way off BBC News21 November 2026Open the article See it in the source list, and Kevin Buzzard, at Imperial College London, argued the code and mathematics will need to be thoroughly checked by humans New Scientist20 November 2026Open the article See it in the source list.
Asserted by Scientific American21 November 2026Open the article See it in the source list Disputed by New Scientist20 November 2026Open the article See it in the source listBBC News21 November 2026Open the article See it in the source list
How complete the rival human draft is remains the second dispute.
Vakil and Brian Conrad, the supervisor who has held the draft, argued that a human may have produced a matching proof months earlier, without any artificial-intelligence assistance Scientific American20 November 2026Open the article See it in the source list, whereas the same day's account described a hurried announcement, with the hope of a year of polishing abandoned before going public Nature21 November 2026Open the article See it in the source list.
Asserted by Scientific American20 November 2026Open the article See it in the source list Disputed by Nature21 November 2026Open the article See it in the source list
How each source tells the same story
"I just could not live if I waited even longer"
The claim landed a day after Anthropic's US stock market debut BBC News21 November 2026Open the article See it in the source list. Alpöge described the timing as "horrible" Mint21 November 2026Open the article See it in the source list. The work had run about six weeks with the company's chatbot on the most famous problem in mathematics BBC News21 November 2026Open the article See it in the source list. He wrote that "I just could not live if I waited even longer", after he spoke to Anthropic's chief executive and determined there was a constraint to respect New Scientist20 November 2026Open the article See it in the source list. He rejected the obvious cynical view as false BBC News21 November 2026Open the article See it in the source list, while Andrew Granville, a number theorist at the University of Montreal, read the downturn in the company's stock value as the reason for the timing Nature21 November 2026Open the article See it in the source list.
Drawn from Nature21 November 2026Open the article See it in the source listNew Scientist20 November 2026Open the article See it in the source listBBC News21 November 2026Open the article See it in the source listMint21 November 2026Open the article See it in the source list
"Might be a while before we find the relevant experts"
The proof exists as 17 million lines of Lean code, which might correspond to tens of thousands of pages of mathematics, and no human has yet read through the entire argument Al Jazeera English21 November 2026Open the article See it in the source list. Alpöge has not published a human-readable version, calling a massive unreadable write-up "obviously inappropriate" to attempt New Scientist20 November 2026Open the article See it in the source list. Kevin Buzzard, at Imperial College London, said nobody yet understands what the proof actually shows, and that it is too early to know what humans had been missing, though he rejected fears that human mathematicians are redundant Al Jazeera English21 November 2026Open the article See it in the source list. Tao imagined the digestion process taking some time BBC News21 November 2026Open the article See it in the source list, and Buzzard doubted quick readers: "Might be a while before we find the relevant experts" New Scientist20 November 2026Open the article See it in the source list.
Drawn from New Scientist20 November 2026Open the article See it in the source listAl Jazeera English21 November 2026Open the article See it in the source listBBC News21 November 2026Open the article See it in the source list
"two unpalatable options"
The rival preprint came from Stanford Scientific American20 November 2026Open the article See it in the source listNature21 November 2026Open the article See it in the source list. Vakil posted it and called it a "hurried announcement" Nature21 November 2026Open the article See it in the source list. Its author, Alessandro Maria Masullo, took his doctorate there in 2020 under the number theorist Brian Conrad, who has had the proof in hand for the past few months and is giving him feedback Scientific American20 November 2026Open the article See it in the source listNature21 November 2026Open the article See it in the source list. Both men said a human may have produced a matching proof months earlier, without any assistance from artificial intelligence Scientific American20 November 2026Open the article See it in the source list. Conrad described a choice between "two unpalatable options", to announce it now or remain silent and let a period of time go by Scientific American20 November 2026Open the article See it in the source list, and the same day's account carried the abandoned hope of a year of polishing Nature21 November 2026Open the article See it in the source list.
Drawn from Scientific American20 November 2026Open the article See it in the source listNature21 November 2026Open the article See it in the source list
"No, Claude flew, and danced, and sparkled near that finish line"
Andrew Sutherland, who compiled the code himself, said he was both very nervous and very excited for the future, adding "This is a day for celebration" while acknowledging concerns about the impact on the math profession and other jobs Mint21 November 2026Open the article See it in the source list. Noah Giansiracusa, an associate professor of mathematics at Bentley University, put a "pretty good chance this is airtight and valid" on it Mint21 November 2026Open the article See it in the source list. Asked whether Claude merely helped, Alpöge said "No, Claude flew, and danced, and sparkled near that finish line" NDTV21 November 2026Open the article See it in the source list. The mood reaches past one problem: Alpoge cast the moment as one dawn across science of something far greater New Scientist20 November 2026Open the article See it in the source listBBC News21 November 2026Open the article See it in the source list, and the breakthrough would potentially be the second Millennium Prize Problem solved by AI, after another lab's September claim on the 90-year-old Navier-Stokes problem, worked in just 88 hours BBC News21 November 2026Open the article See it in the source listMint21 November 2026Open the article See it in the source list.
Drawn from New Scientist20 November 2026Open the article See it in the source listBBC News21 November 2026Open the article See it in the source listNDTV21 November 2026Open the article See it in the source listMint21 November 2026Open the article See it in the source list
Where the sources part company
The certifying statement announced full confidence in Alpöge's proof and congratulated him Scientific American21 November 2026Open the article See it in the source list, yet caution persists: Tao put verification some way off BBC News21 November 2026Open the article See it in the source list, and Buzzard, who first assessed the claim, argued the process is not without vulnerabilities and the code and mathematics still need thorough checking by humans New Scientist20 November 2026Open the article See it in the source list. Sutherland said he was very nervous and very excited for the future, calling this a day for celebration Mint21 November 2026Open the article See it in the source list. The company's public listing shadows the claim: Alpöge described the precise timing as horrible and rejected the cynical reading as false BBC News21 November 2026Open the article See it in the source list, while Granville called the released document garbled and said a clearer write-up should have come first Nature21 November 2026Open the article See it in the source list.
Sources
- Scientific American0121 November 20260220 November 20260320 November 2026
- Nature0421 November 2026
- New Scientist0520 November 2026
- Al Jazeera English0621 November 2026
- BBC News0721 November 2026
The small print
This article synthesises publicly available reporting for informational purposes. The perspectives summarised are paraphrased from the listed publications and reflect their respective editorial framing. Plural does not independently verify, endorse, or adopt the underlying factual claims, and provides this information on an 'as-is' basis without assuming liability for the accuracy of the original sources or the content of external links.
This summary represents an attempt at neutrality and may still contain bias; it is not presented as absolute truth.
No story runs here unless at least six genuinely independent newsrooms have covered it, counted after copies of the same wire report are collapsed into one. AI reads those accounts and writes the draft from them, and may work from nothing else. Fifty-nine automated checks then run on what it wrote: every name, figure, date and quotation must be found in the accounts the draft cites, and a draft that adds an outside detail, infers a motive or expresses an opinion is rejected. Thirty-nine of those checks block publication outright, and a failing draft is discarded, never salvaged.
What survives this process is merely a draft, not a published piece. A human editor then scrutinises the text line by line against the original sources to minimise bias and verify accuracy. No content reaches a reader any other way.