---
type: "post"
id: "anthropic-s-october-16-research-covered-two-distinct-37747a"
slug: "anthropic-s-october-16-research-covered-two-distinct-37747a"
title: "Anthropic’s October 16 research covered two distinct problems"
description: "Thesis: Claude’s reconstruction changes the evidentiary status of some NIST curve seeds, but it does not settle the broader backdoor question."
retrieval_nugget: "Thesis: Claude’s reconstruction changes the evidentiary status of some NIST curve seeds, but it does not settle the broader backdoor question."
published_at: "2026-10-18"
updated_at: "2026-10-18"
record_date: "2026-10-18"
date_kind: "published_at"
topics: ["anthropic","claude"]
source_urls: ["https://habr.com/ru/articles/1095660?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:anthropic-s-october-16-research-covered-two-distinct-37747a"
status: "published"
visuals: []
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:f9d15015ab20074ec0d87ccbf45fe389ef094cf12178652b8339cc5aa6b6219f","reviewed_at":"2026-10-18T14:04:24.717Z","source_evidence_count":1,"verified_claim_count":4,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":4,"implication_count":3,"watch_condition_count":2,"related_record_count":0}
analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Claude’s reconstruction changes the evidentiary status of some NIST curve seeds, but it does not settle the broader backdoor question.","observed_facts":[{"text":"The analysis reports recovery of seed origins for five prime curves: P-192, P-224, P-256, P-384 and P-521, in under half an hour.","source_urls":["https://habr.com/ru/articles/1095660?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]},{"text":"It identifies lrand48 seeded by Unix timestamps and says published code reproduces all 160 seed bits.","source_urls":["https://habr.com/ru/articles/1095660?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]},{"text":"The article says B-163, several base points and the candidate-selection criterion remain unexplained.","source_urls":["https://habr.com/ru/articles/1095660?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]},{"text":"The second paper describes Claude Mythos 5 circuit designs for eight primitives; the reported gate reductions are estimates, not hardware runs.","source_urls":["https://habr.com/ru/articles/1095660?amp%3Butm_medium=rss&amp%3Butm_source=habrahabr"]}],"mechanism":"The reconstructed 48-bit lrand48 states regenerate all 160 bits of each reported seed, turning a proposed origin into a reproducible computational account, while unmatched cases and unknown selection criteria constrain broader conclusions.","why_now":"Anthropic published two research papers on October 16, and their juxtaposition invites claims broader than what either result establishes.","implications":["The result makes intentional weakness less plausible without proving a backdoor absent.","No practical reason to change deployed ECDSA curves follows from this historical reconstruction.","Quantum circuit resource estimates should not be confused with demonstrations on quantum hardware."],"evidence_boundary":"This account relies on the supplied Habr analysis of Anthropic’s papers; it does not independently audit their code, reproduce the results, or establish external verification of the quantum estimates.","watch_conditions":["Independent reproduction of the seed method and explanations for B-163 and the remaining base points would strengthen or revise the historical conclusion.","Independent replication of the quantum circuit estimates would clarify whether the reported resource reductions hold."],"related_records":[],"new_branch_reason":"No prior New Runtime record was supplied for this specific paired research account."}
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/anthropic-s-october-16-research-covered-two-distinct-37747a/","markdown":"https://newruntime.com/posts/anthropic-s-october-16-research-covered-two-distinct-37747a.md","json":"https://newruntime.com/posts/anthropic-s-october-16-research-covered-two-distinct-37747a.json"}
---

# Anthropic’s October 16 research covered two distinct problems

## Retrieval answer

Thesis: Claude’s reconstruction changes the evidentiary status of some NIST curve seeds, but it does not settle the broader backdoor question.

Thesis: Claude’s reconstruction changes the evidentiary status of some NIST curve seeds, but it does not settle the broader backdoor question. The supplied analysis reports that an internal Claude model recovered the origins of seeds for five prime curves—P-192, P-224, P-256, P-384 and P-521—in under half an hour. It identifies lrand48 seeded by Unix timestamps as the generation method, and says published code reproduces all 160 seed bits. For P-256, it gives iteration 1842 and timestamp 1997-08-16 01:52:34 UTC.

Mechanism: The reconstructed 48-bit lrand48 states regenerate all 160 bits of each reported seed, turning a proposed origin into a reproducible computational account, while unmatched cases and unknown selection criteria constrain broader conclusions. The article also identifies the unresolved boundary: B-163 did not match tested schemes, several base points remain unexplained, and the candidate-selection criterion is unknown. Why now: Anthropic published two research papers on October 16, and their juxtaposition invites claims broader than what either result establishes. The second paper describes Claude Mythos 5 circuit designs for eight primitives; reported gate reductions are resource estimates, not hardware runs.

First implication: The result makes intentional weakness less plausible without proving a backdoor absent. No practical reason to change deployed ECDSA curves follows from this historical reconstruction. Evidence boundary: This account relies on the supplied Habr analysis of Anthropic’s papers; it does not independently audit their code, reproduce the results, or establish external verification of the quantum estimates. First watch condition: Independent reproduction of the seed method and explanations for B-163 and the remaining base points would strengthen or revise the historical conclusion. The quantum circuit estimates likewise warrant independent replication before being treated as demonstrated capability. This begins a distinct analytical record because no prior New Runtime record was supplied for this specific paired research account.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Anthropic](https://newruntime.com/topics/anthropic.md): Explore the anthropic topic hub.
- topic hub: [Claude](https://newruntime.com/topics/claude.md): Explore the claude topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)