Anthropic Mathematician Claims AI Helped Prove Riemann Hypothesis
Levent Alpoge states that Anthropic's Claude AI assisted in solving the 167-year-old mathematical problem just one day after the company went public.
By Amir Navarro ·

Levent Alpoge, a Turkish American mathematician employed by the San Francisco-based artificial intelligence firm Anthropic, has asserted that the company's software, Claude, has assisted in proving the Riemann hypothesis. This claim suggests a significant milestone in the application of AI to complex mathematical challenges. Alpoge published a note on his personal website on Friday, detailing how the AI system contributed to developing a solution over a period of approximately six weeks.
The Riemann hypothesis, proposed by German mathematician Bernhard Riemann in 1859, is one of the most famous unsolved problems in mathematics. It concerns the distribution of prime numbers along the number line. The problem is one of seven Millennium Prize problems selected by the Clay Mathematics Institute in 2000, each carrying a prize of one million dollars for a verified solution.
Alpoge stated that while he and Claude had collaborated extensively in recent months, the AI significantly accelerated the process near the end. He described the final stages as a period where the software outpaced him, writing that Claude flew, danced, and sparkled near the finish line. To support his claim, he provided a link to computer code written in the formal language Lean, which he said could verify the findings.
The timing of the announcement has drawn attention because it occurred just one day after Anthropic went public on the Nasdaq stock exchange. Alpoge acknowledged the awkwardness of the timing, noting that he would have preferred to complete the work earlier to avoid the appearance of coincidence with the company's market debut.
The mathematical community has begun the process of verifying Alpoge's proof. This scrutiny comes amid broader discussions about the role of private technology companies in advancing fundamental mathematics. Some mathematicians have expressed unease about the financial scale and methods employed by lucrative AI firms to solve long-standing problems.
In September, Anthropic's competitor OpenAI announced that a non-public version of its AI tools had helped solve part of another Millennium Prize problem. That effort involved a private company valued at nearly a trillion dollars spending fifteen million dollars to employ ten thousand AI systems. The approach struck many observers as unusual, with critics questioning the necessity of such a massive financial and computational investment for a single mathematical task.
Alpoge was working on that specific problem for Anthropic when OpenAI made its announcement. The situation has fueled debate within the academic community regarding the reliance on artificial intelligence for discovery. University of Warwick mathematician James Robinson told the Guardian in September that it is frustrating to see large tech companies using vast resources to demonstrate the capabilities of their latest models.
Robinson characterized the behavior as immature playground boasting, supported by billions of dollars and the potential for significant environmental damage. He noted that this occurs in an era where climate change is considered a major global challenge. The debate highlights the tension between the potential for AI to accelerate mathematical progress and the concerns over the resource-intensive methods used by private corporations.
Meanwhile, another independent mathematician, Alessandro Masullo, released an incomplete proof of the Riemann hypothesis on the same Friday. Masullo stated that he had been working on the problem for almost twelve years and chose not to use AI because he did not fully trust the system. His announcement adds another layer to the ongoing discussion about how future mathematical breakthroughs might be achieved.
As mathematicians work to confirm or refute Alpoge's claim, the episode underscores the shifting landscape of mathematical research. The integration of AI tools like Claude into the workflow of professional mathematicians raises questions about the nature of discovery and the balance between human intuition and computational power in solving problems that have resisted solution for nearly two centuries.






