🔍 Read the full analysis: A New Era For The Cryptography That Supports Finance on ThorstenMeyerAI.com
Get privacy and security gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
OpenAI published 722 AI-generated mathematical manuscripts on October 6, while computer scientists and cryptocurrency figures raised questions about how AI could affect cryptographic assumptions. No cryptographic protocol has been shown to be broken, and the claims require independent verification. The development complicates planning for finance and other sectors already preparing for quantum computers.
OpenAI published 722 mathematical manuscripts on October 6, and the release has prompted renewed discussion about whether AI could find algorithms that weaken cryptography used in finance and other sectors. No cryptographic system has been shown to be broken; the concern is that mathematical assumptions underpinning security may be tested in new ways, alongside the established risk from quantum computers.
The manuscripts were generated by an unreleased internal model across 372 families of work, according to the source report. The model worked on roughly 4,000 problems, with the report saying each result used an average of about three hours of ChatGPT Pro compute. The reported topics include the Unique Games Conjecture, Hilbert’s tenth problem over the rationals and a zero-free region for the Riemann zeta function. These are presented as claims requiring expert checking, not as settled mathematical results.
Some work attracting attention concerns computational complexity: how quickly particular problems can be solved. The report points to results involving integer multiplication and Fourier transforms, and a separate paper by Virginia Vassilevska Williams and Josh Alman that describes a faster-than-expected approach to 3SUM. It says an Anthropic model contributed a key idea to that work. Such findings can matter to cryptography if they change the expected difficulty of underlying mathematical problems, but a faster algorithm for one task does not itself demonstrate a practical attack on a deployed security system.
Computer scientist Scott Aaronson noted that cryptography was absent from the 722 published manuscripts, according to the report. Aaronson also said his sources indicated that AI companies were discreetly testing whether internal models could break important protocols. Those tests and their results have not been publicly documented in the supplied material. The source also reports that OpenAI withdrew a claimed proof concerning the Hodge conjecture for products of K3 surfaces after a sign error was identified, illustrating why independent verification matters.
The old map is gone: AI mathematics, quantum computers and the cryptography holding up finance and defence
For a decade the plan was simple: elliptic curves doomed by quantum; lattices safe; hashes safe. Nothing has been broken. But a second threat has arrived that doesn’t respect those borders — AI producing new mathematics faster than any human community, against assumptions that are believed, not proven.
Now: on borrowed time — possibly shorter than the quantum countdown suggests.
Now: unproven against AI — and the destination most of the world is migrating to.
Now: reminded estimates move — BSI advised against new deployments on 1 Oct 2026.
Now: safest ground available — not a guarantee.
~n log0.9999999999999 n — a barrier many thought fundamental (OpenAI, claimed)
Overturns a half-century conjecture. Williams & Alman; key idea from an Anthropic model
“Conspicuous by its absence” (Aaronson) — labs reportedly testing crypto “gingerly and discreetly”
ECDSA could break before Q-day, “in the worst case in months not years.” Move funds to never-signed addresses. ~6M BTC sit behind exposed keys.
The new risk is the destination of the migration. Hash-only where possible; “much more paranoid” lattice params; ×10 key sizes long-term. Doesn’t recommend anyone scramble.
“No evidence whatsoever” that elliptic-curve assumptions are close to failing.
Classical breaks could reach “quantum-safe” schemes — but don’t treat a two-year scenario as a date.
Known to IBM and the NSA designing DES (~1974); public via Biham & Shamir (~1990); confirmed by Coppersmith (1994).
Invented at GCHQ — RSA- and Diffie–Hellman-equivalents — and kept secret for over two decades.
No crypto in 722 manuscripts. Found and withheld? Not posed? Posed and failed? Indistinguishable from outside.
Traffic recorded today is decrypted when a break arrives. For secrets that must last 25+ years, a break in 2035 is a break today. A state that finds one won’t announce it — it will mine its archives.
Signatures can be built from hashes. Encryption and key exchange need a trapdoor with structure — lattices, codes or group theory. Defence can only choose which structure, how much margin, how many combined.
Every date was set against quantum hardware forecasts with visible warning. The AI threat offers none.
“ML-KEM everywhere” means starting over if lattices weaken. “We can swap algorithms” doesn’t.
Blockchains show a classical break first — exposed keys and balances are public. Monitor dormant exposed addresses.
Every algorithm, key, certificate, protocol.
PQ + classical, as BSI requires.
Firmware, updates, long-term keys.
Highest sets; evaluate FrodoKEM.
More than one mathematical family; HQC coming.
Swap algorithms without rebuilding.
Forward secrecy, rotation, hidden keys.
Buterin: lost more in botched migrations than in all hacks.
Nothing has been broken, and the sceptics are right that there’s no evidence elliptic curves or lattices are about to fall. But the map has changed: elliptic curves on borrowed time, lattices unproven against AI, codes reminded that estimates move, hashes the safest ground available. For finance, intelligence and defence the answer is the same whichever threat arrives first.The quantum threat comes with a countdown. The AI threat may arrive as a silence — an empty folder where a paper should have been. The winners will be those who can change their algorithms fastest.
Security Planning Beyond Quantum Risk
Financial institutions use cryptography to protect communications, authenticate transactions and secure digital assets. Governments, intelligence services and defence organisations also rely on it. The immediate significance is not evidence of a breach, but a possible change in how security teams assess the mathematical assumptions behind systems that are meant to remain secure for years.
Quantum risk has a recognizable hardware component: a sufficiently capable, error-corrected quantum computer could use Shor’s algorithm against RSA and elliptic-curve cryptography. An AI-assisted mathematical discovery might instead be an algorithm running on conventional computers. If one were found, its existence might not be apparent until it was used or disclosed. That makes monitoring and response planning harder, though the scale and timing of any such risk are unknown.
The distinction matters for finance because cryptographic changes can involve large systems, long-lived records and coordinated upgrades across institutions. A prudent response is to follow verifiable research and review migration plans, not to treat speculative warnings as proof that current systems have failed. No evidence in the source establishes a practical AI-derived cryptographic attack.
quantum-resistant cryptography hardware
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The Existing Post-Quantum Migration
For years, security planning has treated large-scale quantum computing as a future threat to public-key systems including RSA and elliptic-curve cryptography. In August 2024, the U.S. National Institute of Standards and Technology standardized three post-quantum algorithms: ML-KEM for establishing encryption keys, ML-DSA for digital signatures and SLH-DSA, a hash-based signature scheme. Organizations have been working to assess and adopt replacements.
The source report argues that AI-generated mathematical work could complicate that plan because lattice-based cryptography, including the basis for ML-KEM and ML-DSA, also depends on assumptions about computational difficulty. That is a concern to investigate, not a finding that the standards are unsafe. The report contrasts lattices with hash-based approaches, but it provides no demonstrated attack against either category.
Cryptocurrency has made the debate especially visible because some blockchains expose public keys after transactions are signed. On October 7, Ethereum Foundation researcher Justin Drake advised the industry to plan calmly for a possible “bunker mode,” including moving funds to addresses whose public keys have not been exposed. He warned that an elliptic-curve break could, in a worst case, arrive before a quantum computer. The report also cites an estimate of roughly six million bitcoin in addresses with exposed public keys; it supplies no independent methodology for that figure.
Ethereum co-founder Vitalik Buterin cautioned against rushing to move funds, while raising concern about post-quantum lattices and other mathematical systems. His comments frame the issue as a question about whether algorithms could reveal unexpected weaknesses—not as confirmation that such weaknesses have been found.
As an affiliate, we earn on qualifying purchases.
No Demonstrated Cryptographic Break
The central unknown is whether AI systems can produce a practically useful attack on a cryptographic protocol, and whether any such result has already been found privately. The supplied report describes discreet testing based on sources, but gives no named test, technical result or independent confirmation. It also does not establish that the OpenAI manuscripts contain a cryptographic attack.
The mathematical claims themselves remain subject to verification. The reported correction to the Hodge-conjecture proof is a reminder that generated proofs may contain errors. It is also unclear whether any faster computational methods described in related work can be extended to problems on which real-world cryptography relies, or whether they would provide a meaningful advantage at practical key sizes.
There is no confirmed timeline for an AI-related threat, no public evidence that existing keys have been recovered, and no basis here for concluding that NIST’s post-quantum standards are compromised. The reported bitcoin estimate and Drake’s worst-case timeframe should be treated as attributed claims, not established forecasts.
As an affiliate, we earn on qualifying purchases.
Verification and Migration Decisions
The next step is independent mathematical review of the AI-generated manuscripts and any reported complexity results. For cryptography, the key milestone would be a reproducible, technically described result showing that an established security assumption can be weakened—and an assessment of whether it works at realistic scales. Until then, the source supports continued scrutiny, not a declaration that a break has occurred.
Financial institutions and public agencies will need to continue their post-quantum inventories and migration work while tracking credible research on both quantum and classical algorithms. Blockchain developers and users may debate protective measures for exposed public keys, but the cited comments do not amount to a consensus recommendation for immediate wallet transfers. Further public evidence and expert validation will determine whether the discussion changes standards, migration schedules or operational guidance.
As an affiliate, we earn on qualifying purchases.
Key Questions
Has AI broken a financial or cryptocurrency cryptographic system?
No break is confirmed in the supplied source. It describes concern about possible algorithmic discoveries and reports that companies may be testing models, but gives no verified attack or recovered key.
What did OpenAI publish?
OpenAI published 722 mathematical manuscripts across 372 families on October 6, according to the source report. Their claims require checking, and the release is not itself evidence of a cryptographic vulnerability.
How is the AI concern different from quantum risk?
A quantum attack would depend on building sufficiently capable hardware. The concern discussed here is that an AI system might help discover a better algorithm that can run on ordinary computers. Whether that can happen in a practically useful way remains unknown.
Are post-quantum cryptography standards known to be unsafe?
No. The source raises questions about the mathematical assumptions behind lattice-based standards such as ML-KEM and ML-DSA, but reports no demonstrated attack against them.
Should cryptocurrency holders move funds now?
The source does not establish a need for an immediate move. Justin Drake called for calm planning, while Vitalik Buterin explicitly advised against scrambling to move funds that day. Readers should distinguish those attributed views from confirmed evidence of an active attack.
Source: ThorstenMeyerAI.com
Columbus Day / Indigenous Peoples' Day Picks
long weekend sales
As an affiliate, we earn on qualifying purchases.
