Claude solves decades-old mystery behind NSA cryptographic numbers

Anthropic says its Claude AI has reconstructed the origins of unexplained numbers used in widely adopted security standards, tracing them to a conventional random-number generator initialized with dates and times from the 1990s. Bill Mann wrote cyberinsider.com's report on October 17, 2026. The discovery helps answer whether the U.S. National Security Agency deliberately chose those numbers to introduce hidden weaknesses. In a report published October 16, Anthropic said an internal Claude research model identified how the seeds for five standardized mathematical curves were generated in under half an hour.

This piece covers the generator Claude recovered, the timestamps on the five curves, the independent check, the bounty, and the gaps that remain. It is for engineers who implement the Elliptic Curve Digital Signature Algorithm and for readers following the debate over National Security Agency seeds in standards used to verify authenticity and detect tampering.

Claude solves decades-old mystery behind NSA: what actually changed

The curves the U.S. National Institute of Standards and Technology (NIST) published in the 2000 digital signature standard underpin ECDSA, and the NSA supplied the seeds used to derive them. Their origins were unexplained. Researchers could check whether a seed produced particular mathematical parameters, but could not explain why that starting number was chosen. That gap fueled speculation about possible backdoors. Earlier investigations followed an incorrect lead suggesting the numbers came from humorous phrases. Anthropic's October 16 report replaces that lead.

For the five NIST curves, the recovered timestamps fall between May 1997 and April 1998. The P-256 seed was reproduced using August 16, 1997, at 01:52:34 UTC and the 1,842nd candidate generated. Anthropic published code allowing others to reproduce the results. NIST published the parameters in the 2000 standard. ECDSA implementations still use the five NIST curves to verify authenticity and detect tampering. An internal Claude research model identified the origin of the NSA seeds in under half an hour.

Claude solves decades-old mystery behind NSA: how it works

According to Anthropic, Claude approached the mystery by considering how software from that era would have generated the numbers. It tested recognizable patterns left by older random-number generators, then used mathematical techniques to reconstruct their internal state. The results pointed to lrand48, a generator available in C programming libraries, initialized with Unix timestamps, numbers representing dates and times. A predictable generator would be dangerous if used to create private keys. Here, however, it produced starting material for mathematical parameters that were always public.

The P-256 case is the fully specified example: August 16, 1997, at 01:52:34 UTC, and the 1,842nd candidate generated. That timestamp initialized lrand48. Anthropic published code allowing others to reproduce the results. Cryptographer Filippo Valsorda, who organized a bounty for solving the mystery, independently reimplemented the generation algorithm in Go and verified the finding. His verification supports the mathematical result. The account of Claude's discovery process and timing comes from Anthropic, including the under-half-an-hour run by the internal research model.

Claude solves decades-old mystery behind NSA: why it matters now

Unexplained seeds kept open the question of hidden weaknesses: whether the NSA deliberately chose the numbers later published by NIST. Researchers could check a seed against the parameters but not why that starting number was chosen. That gap fueled speculation about possible backdoors. Anthropic says the findings instead strengthen confidence that the curves were selected without hidden weaknesses. The origin attached to the five curves is lrand48 initialized with Unix timestamps from May 1997 to April 1998, a conventional generator from C libraries.

Reconstructing their origins does not expose users' private keys, enable forged signatures, or unlock encrypted communications. A predictable generator would be dangerous if used to create private keys. Here, however, it produced starting material for mathematical parameters that were always public. Lukasz Olejnik wrote on October 17, 2026 that the NSA "seeded its random numbers like a first-year student," and that Claude tested the decades-old theory of some cryptographers that the NSA secretly planted weak NIST elliptic curves behind unexplained seeds.

Claude solves decades-old mystery behind NSA: who is affected

Implementers of ECDSA ship these curves so computers can verify authenticity and detect tampering. NIST published them in the digital signature standard in 2000, using seeds the NSA supplied, so an unexplained seed became a question for widely adopted security standards. Anthropic says the findings strengthen confidence that the curves were selected without hidden weaknesses. Reconstructing the origins does not expose private keys, enable forged signatures, or unlock encrypted communications, because the generator produced public starting material rather than secret keys.

Filippo Valsorda organized the bounty, independently reimplemented the generation algorithm in Go, and verified the finding. He confirmed that the Anthropic team claimed the prize and chose a charitable payout of $36,864 to the Internet Security Research Group, which operates Let's Encrypt. Bill Mann's October 17, 2026 article pairs Anthropic's process description with Valsorda's confirmation. Researchers who still lack an explanation are also affected: skipped candidate curves, one seed from a separate family, and several standardized starting points remain open.

Claude solves decades-old mystery behind NSA: what to watch

Some details remain unresolved. The generation process skipped candidate curves that appear to meet the relevant requirements, and the researchers do not know why. They also could not explain one seed from a separate family of curves or the origins of several standardized starting points used in signature calculations. P-256 is already pinned to August 16, 1997, at 01:52:34 UTC and the 1,842nd candidate, in the May 1997 to April 1998 window. Anthropic published code allowing others to reproduce the results.

The discovery substantially narrows the uncertainty surrounding these standards, leaving only a few gaps in their history. Filippo Valsorda has verified the generation result in Go and confirmed the $36,864 payout to the Internet Security Research Group, which operates Let's Encrypt. He confirmed that the Anthropic team claimed the bounty he had organized and chose that charitable payout. The account of Claude's discovery process and timing, including the internal research model finishing in under half an hour, comes from Anthropic.