# Riemann hypothesis implications: sharper prime counts, no broken RSA

> A proof would cap the prime-counting error at about √x log x and confirm many conditional theorems, but RSA rests on factoring, which the hypothesis does not touch.

URL: https://claudegpt.si/news/riemann-hypothesis-implications  
Published: 2026-11-21T18:59:38.854Z  
Author: ClaudeGPT Si Newsroom  
Section: Research  
Topics: Riemann hypothesis, Mathematics, Lean, Formal verification, Claude, Anthropic  
Publisher: ClaudeGPT Si

## Key takeaways

- The Riemann hypothesis is equivalent to the prime count π(x) staying within about √x log x of Li(x), far tighter than the error bound proven today.
- A proof of RH and GRH would turn many conditional results, such as Miller's deterministic primality test and Robin's divisor inequality, into theorems.
- It would not break RSA, whose security rests on the difficulty of factoring, and the Clay prize needs journal publication plus two years of general acceptance.

The Riemann hypothesis implications are mostly inside mathematics. A proof would guarantee that the count of primes up to any number x stays within roughly √x times log x of its predicted value, and it would turn many results proved "assuming the Riemann hypothesis" into theorems. It would not break RSA or other prime-based encryption, because that security rests on how hard it is to factor large numbers, and the hypothesis is not about factoring.

The question is live because on November 20, 2026, Anthropic mathematician Levent Alpöge said he had proved the hypothesis with Claude, and on November 21 the Lean and Mathlib communities [said they are fully confident](https://leanprover-community.github.io/blog/posts/riemann-hypothesis/) that his computer-checked proof is correct. No human-readable version exists yet.

## What the Riemann hypothesis says

The Riemann zeta function starts as the sum 1 + 1/2^s + 1/3^s and so on, extended by Riemann to almost every complex number s. Apart from the "trivial" zeros at -2, -4, -6 and so on, the [official Clay problem description](https://www.claymath.org/wp-content/uploads/2022/05/riemann.pdf) by E. Bombieri states the hypothesis in one line: the nontrivial zeros have real part 1/2. In simple terms, those zeros control how the primes are spread out, and the hypothesis puts them all on one line, the arrangement that makes the primes as regular as possible. Our explainer on [what the Riemann hypothesis is](https://claudegpt.si/news/what-is-the-riemann-hypothesis) has the background.

## Riemann hypothesis and prime numbers: the error shrinks to about √x

Write π(x) for the number of primes up to x. The prime number theorem, proved in 1896, says π(x) is close to a smooth function, Li(x). The question is how close.

| | Error bound for π(x) versus Li(x) | Source |
| --- | --- | --- |
| Proven today | about x · exp(-0.2098 (log x)^(3/5) / (log log x)^(1/5)) | Ford, 2002 |
| If RH is true | about √x · log x | von Koch, 1901 |
| If RH is true, explicit | under √x · log x / (8π), for x of at least 2,657 | Schoenfeld, 1976 |

The proven bound, per [Wikipedia's prime number theorem article](https://en.wikipedia.org/wiki/Prime_number_theorem), is only slightly smaller than x itself. Von Koch showed in 1901 that the √x log x bound is equivalent to the hypothesis. For x = 10^12, Schoenfeld's formula would guarantee the count of primes below a trillion to within about 1.1 million. Bombieri notes the error can't be pushed much lower. If the hypothesis failed, he writes, the distribution of primes would be thrown into havoc.

A proof leaves some questions open. Assuming RH, Cramér showed the gap after a prime p is at most on the order of √p log p, while his own conjecture puts gaps near (log p)^2, according to [Wikipedia's Riemann hypothesis article](https://en.wikipedia.org/wiki/Riemann_hypothesis).

## Theorems proved conditionally on RH and GRH

The generalized Riemann hypothesis (GRH) makes the same claim for Dirichlet L-functions, which count primes in sequences such as 3, 7, 11, 15. Our piece on [RH vs GRH](https://claudegpt.si/news/riemann-hypothesis-vs-generalized-riemann-hypothesis) explains the difference. Results that would become unconditional include:

- **Robin's inequality**: the divisor sum of n is below e^γ · n · log log n for every n above 5,040. Robin showed in 1984 that this is equivalent to RH.
- **Miller's primality test**: G. Miller showed in 1976 that, under a generalized hypothesis, testing bases up to about 2(ln n)^2 decides primality in polynomial time ([Wikipedia](https://en.wikipedia.org/wiki/Miller%E2%80%93Rabin_primality_test)). Bombieri lists it among the consequences.
- **Weak Goldbach**: Hardy and Littlewood showed in 1923 that GRH implies every large enough odd number is a sum of three primes.
- **Number fields**: Bombieri lists a very strong effective form of Chebotarev's density theorem.
- **Landau-Siegel**: Terence Tao noted on [his November 20 post](https://terrytao.wordpress.com/2026/11/20/unforced-euler-singularity-unforced-navier-stokes-singularity-and-the-generalized-riemann-hypothesis/) that its resolution follows from GRH.

Several were later proved without the hypothesis: a polynomial-time primality test by Agrawal, Kayal and Saxena in 2002, and weak Goldbach by Vinogradov (1937) and Helfgott (2013). Bombieri counts such cases as evidence the hypothesis is true.

Scope matters. Alpöge's Lean proof covers RH and GRH for Dirichlet L-functions, according to [a verification page](https://math.mit.edu/~drew/grh/) kept by MIT's Andrew Sutherland. Dedekind zeta functions and broader L-functions were not among the statements checked, so results that need those wider versions are not covered.

## Riemann hypothesis and cryptography: why RSA would not break

RSA uses two large secret primes. The [1978 RSA paper](https://people.csail.mit.edu/rivest/Rsapaper.pdf) by Rivest, Shamir and Adleman says its security "rests in part on the difficulty of factoring the published divisor, n," the product of those primes. Whoever factors n can compute the private key. The Riemann hypothesis describes how many primes there are and how evenly they are spread. It gives no method for finding the two primes inside a given n.

The paper also says nobody has proved factoring is hard. RSA is trusted because centuries of effort have not produced a fast method. A proof of the hypothesis doesn't change that either way.

Key generation doesn't need the hypothesis. The paper's method: pick random 100-digit odd numbers until one is prime. The prime number theorem, proven without RH, says about 115 tries will do, and a probabilistic test leaves a one-in-2^100 chance of a composite slipping through. GRH would only guarantee that Miller's deterministic test is correct, and a deterministic polynomial-time test has existed without any hypothesis since 2002.

Scientific American's Lee Billings [made the same point](https://www.scientificamerican.com/article/what-the-riemann-hypothesis-proof-really-means-about-ai-and-the-future/) on November 20: testing primality differs from factoring, and GRH offers no known shortcut for factoring. Nobody has yet read the claimed proof as mathematics, and nothing published about it mentions factoring.

## Clay Millennium Prize: the $1 million conditions

The [prize rules](https://www.claymath.org/wp-content/uploads/2022/03/millennium_prize_rules_0.pdf), approved on September 26, 2018, require:

| Condition | What the rules say |
| --- | --- |
| Publication | In a refereed mathematics publication of worldwide repute, though Clay can relax this |
| Waiting period | At least two years after that publication |
| Acceptance | General acceptance in the global mathematics community, as Clay judges it |
| Review | A Clay advisory committee of at least three, two of them experts |
| Submissions | Clay does not accept solutions sent to it directly |

The prize can be split, and a counterexample can also win. Lean files on a personal website are not a qualifying publication, so the two-year clock has not started, and Clay's [problem page](https://www.claymath.org/millennium/riemann-hypothesis/) still says "Unsolved."

## Where the Alpöge claim stands on November 21

- **The claim.** In a [statement](https://alpo.ge/remark.pdf), Alpöge wrote that he believes he has, "working with Claude, discovered a proof of the Riemann Hypothesis," and that it extends to GRH. He released only Lean code and plans a proper write-up. See our [news story on the claim](https://claudegpt.si/news/riemann-hypothesis-claude-proof-claim).
- **Formal check passed.** The Lean and Mathlib communities, endorsed by the Lean FRO and the Mathlib Initiative, certify that the 17-million-line library proves exactly Mathlib's statement of RH, after several independent proof checkers accepted it. Our article on [the Lean verification](https://claudegpt.si/news/riemann-hypothesis-lean-proof-verified) has more.
- **Still running.** Sutherland's page, as of 12:55 UTC on November 21, reports all 175,187 modules compile using only Lean's standard axioms, with some kernel checks unfinished.
- **Not done.** No mathematician has read a human explanation. Tao expects digesting it to take time.
- **The other claim.** Ravi Vakil [announced](https://math216.wordpress.com/2026/11/20/hurried-announcement-alessandro-masullo-grh/) a separate draft proof of GRH by Alessandro Masullo, made without AI. He later said people had raised possible problems that must be resolved first.

Watch for the remaining checks on Sutherland's page, a written paper from Alpöge, and a journal submission, the step that starts Clay's two-year clock.

## Frequently asked questions

### Would proving the Riemann hypothesis break cryptography?

No. RSA's security rests on the difficulty of factoring the public number n into its two primes, and the Riemann hypothesis describes how primes are distributed on average, not how to find the factors of a given number. Key generation already works without it, using the prime number theorem and probabilistic primality tests.

### What is the Riemann hypothesis in simple terms?

It says every nontrivial zero of the Riemann zeta function lies on one line, the line where the real part is 1/2. Those zeros control how the primes are spread out, so the hypothesis means the primes are as evenly spread as they can be.

### How is the Riemann hypothesis related to prime numbers?

Riemann's 1859 formula counts primes using the zeros of the zeta function. The hypothesis is equivalent to the number of primes up to x differing from Li(x) by at most about √x times log x, as von Koch showed in 1901.

### Has the Riemann hypothesis been proved?

On November 21, 2026, the Lean and Mathlib communities said they are fully confident in Levent Alpöge's computer-checked Lean proof, but no human-readable version has been published and the Clay Mathematics Institute still lists the problem as unsolved.

## Sources

- [Problems of the Millennium: the Riemann Hypothesis (E. Bombieri)](https://www.claymath.org/wp-content/uploads/2022/05/riemann.pdf), Clay Mathematics Institute
- [The Riemann Hypothesis](https://www.claymath.org/millennium/riemann-hypothesis/), Clay Mathematics Institute
- [Millennium Prize Description and Rules](https://www.claymath.org/wp-content/uploads/2022/03/millennium_prize_rules_0.pdf), Clay Mathematics Institute
- [A Method for Obtaining Digital Signatures and Public-Key Cryptosystems](https://people.csail.mit.edu/rivest/Rsapaper.pdf), R. L. Rivest, A. Shamir, L. Adleman (MIT)
- [Lean verification of the Riemann Hypothesis](https://leanprover-community.github.io/blog/posts/riemann-hypothesis/), Lean and Mathlib communities
- [RH Lean proof: verification files](https://math.mit.edu/~drew/grh/), Andrew Sutherland, MIT
- [Statement on the Riemann Hypothesis](https://alpo.ge/remark.pdf), Levent Alpöge
- [Unforced Euler singularity, unforced Navier-Stokes singularity, and the generalized Riemann hypothesis](https://terrytao.wordpress.com/2026/11/20/unforced-euler-singularity-unforced-navier-stokes-singularity-and-the-generalized-riemann-hypothesis/), Terence Tao
- [Hurried announcement: Alessandro Masullo, GRH](https://math216.wordpress.com/2026/11/20/hurried-announcement-alessandro-masullo-grh/), Ravi Vakil
- [What the Riemann hypothesis proof really means about AI and the future](https://www.scientificamerican.com/article/what-the-riemann-hypothesis-proof-really-means-about-ai-and-the-future/), Scientific American
- [Prime number theorem](https://en.wikipedia.org/wiki/Prime_number_theorem), Wikipedia
- [Riemann hypothesis](https://en.wikipedia.org/wiki/Riemann_hypothesis), Wikipedia
- [Miller-Rabin primality test](https://en.wikipedia.org/wiki/Miller%E2%80%93Rabin_primality_test), Wikipedia
