Anthropic model traces NIST curve seeds, easing old backdoor fears but leaving bitcoin quantum risk intact

Anthropic model traces NIST curve seeds, easing old backdoor fears but leaving bitcoin quantum risk intact
AI demystifies NIST seeds

Decades-old questions over how the U.S. government’s standard elliptic curves were created have resurfaced after an Anthropic AI model reconstructed the number-generation process behind several NSA-linked seeds. The finding suggests the disputed values came from a basic legacy random-number program rather than a hidden trap, while leaving bitcoin’s separate exposure to future quantum computing unchanged.

Highlights

  • Anthropic traced NIST elliptic curve seeds to the lrand48 random-number generator using 1997-1998 timestamps, resolving long-standing concerns over NSA backdoors.
  • Bitcoin wallets remain unaffected as they use secp256k1, whose design offers fewer uncertain parameters, but the result highlights risks from weak random number generators.
  • The finding does not eliminate bitcoin’s vulnerability to quantum computing attacks, which could still expose private keys from public keys on the blockchain.

NIST curve mystery revisited

As first reported by CoinDesk, an Anthropic internal research model traced five seeds used in NIST elliptic curves to lrand48, a random-number generator found in older C software libraries, using timestamps from 1997 and 1998.

The result addresses a question that has circulated in cryptography since 2000, when the U.S. National Institute of Standards and Technology adopted curves supplied by the National Security Agency without explaining the origin of some numbers used to generate them. Because those seeds were public but unexplained, some researchers suspected the NSA might have searched through many options to find curves with a weakness known only to the agency.

That suspicion intensified after the Snowden disclosures tied the NSA to broader efforts to weaken cryptographic standards. NIST later removed the Dual EC random-number generator from its recommendations, adding to concern around opaque design choices in government-backed cryptography.

Anthropic’s model reportedly took less than half an hour to reproduce the sequence behind P-256, one of the most widely used curves, finding an exact match for its published seed at the 1,842nd candidate. The company said in a blog post that users of NIST curves do not need to make changes based on the discovery.

Implications for bitcoin and digital security

For bitcoin, the finding narrows one line of speculation but does not materially change the network’s security outlook. Bitcoin wallets use secp256k1 rather than the NIST curves under review, and the article says the bitcoin curve is built from a simpler equation with two small fixed numbers, leaving far less room for adjustable parameters.

The reconstruction also underscores why predictable randomness would be dangerous in wallet generation. If a secret key were created with a weak generator such as lrand48, an attacker who could estimate when it was produced might be able to recreate the key and seize the funds.

That does not remove bitcoin’s longer-term risk from quantum computing. A sufficiently powerful quantum machine could derive a secret key from a public key exposed on the blockchain regardless of how the underlying curve was selected, preserving a separate threat for the crypto sector and other systems that rely on public-key cryptography.

The search carried a $36,864 bounty, which will go to the Internet Security Research Group, the nonprofit behind Let’s Encrypt.

Our earlier report on Cohere’s planned acquisition of AISLE explained how the deal would add AI-driven software vulnerability detection to Cohere’s broader enterprise security push alongside its expansion in Europe. We noted that AISLE’s technology had already identified hundreds of new bugs in widely used software, including multiple OpenSSL vulnerabilities that were subsequently fixed—highlighting how advanced AI is increasingly being used to uncover and mitigate security weaknesses.

This material may contain third-party opinions, none of the data and information on this webpage constitutes investment advice according to our Disclaimer. While we adhere to strict Editorial Integrity, this post may contain references to products from our partners.