🔍 Read the full analysis: Finance And Defence Navigate New Risks In AI And Cryptography on ThorstenMeyerAI.com
Get the latest gadgets 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
A source report says OpenAI published 722 mathematical manuscripts on Oct. 6, including results that challenge some longstanding expectations about computational speed. The work has not established that cryptography is broken, but warnings from cryptocurrency figures have brought attention to the possibility that AI-assisted mathematics could expose weaknesses in both current and post-quantum systems.
OpenAI published 722 mathematical manuscripts on Oct. 6, according to the source report, prompting renewed concern that advances in AI-assisted mathematics could eventually affect the assumptions underpinning cryptography. No cryptographic protocol has been shown to be broken, but Ethereum Foundation researcher Justin Drake and co-founder Vitalik Buterin have publicly discussed risks to both existing systems and post-quantum replacements.
The manuscripts reportedly came from an unreleased internal model, which worked on roughly 4,000 problems and produced results across 372 families. The source says the work included claims concerning the Unique Games Conjecture, Hilbert’s tenth problem over the rationals and a zero-free region for the Riemann zeta function. Those are mathematical claims, not independently established cryptographic breaks.
Some of the attention has focused on computational results. The source cites computer scientist Scott Aaronson’s account of claims involving faster-than-expected integer multiplication and Fourier transforms. It also describes a result for 3SUM, a computational problem, attributed to a paper by Virginia Vassilevska Williams and Josh Alman. The source says an Anthropic model contributed the key idea to that work. These results may challenge expectations about algorithmic limits, but they do not by themselves demonstrate that encryption can be defeated.
The source also reports that OpenAI withdrew a claimed proof related to the Hodge conjecture for products of K3 surfaces after a sign error was identified. That correction illustrates why the manuscripts require checking. The source says AI companies are discreetly testing whether internal models can break important protocols, citing Scott Aaronson’s account of what his sources told him; it provides no public result showing that such a test has succeeded.
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.
Pressure on Security Migration Plans
Financial institutions, intelligence agencies and defence organisations rely on cryptographic systems to protect communications, transactions and stored information. Their current migration planning has largely treated quantum computing as the principal future challenge: a sufficiently capable quantum computer could use Shor’s algorithm against RSA and elliptic-curve cryptography. AI-assisted discovery raises a different concern: an improved algorithm could run on conventional hardware and might not be visible before it is used.
That possibility matters to post-quantum planning. The US National Institute of Standards and Technology issued its first major post-quantum standards in August 2024, including ML-KEM and ML-DSA, which are lattice-based, and SLH-DSA, based on hash functions. The reported research does not show those standards are vulnerable. It does, however, underline that mathematical confidence is not the same as proof that no more efficient attack exists.
For organisations responsible for long-lived or high-value information, the practical issue is not that they should abandon current standards on the basis of unverified claims. It is that security plans may need to account for more than a hardware countdown. A software-based discovery could be difficult to detect, and a useful algorithm might be kept private by whoever found it.
post-quantum cryptography security tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From Quantum Risk to Algorithm Risk
The established quantum concern is relatively specific: a large, error-corrected quantum computer could threaten public-key systems such as RSA and elliptic-curve cryptography. That prospect has led governments and companies to plan migrations to post-quantum cryptography. NIST’s 2024 standards provided widely available replacement algorithms, including lattice-based schemes and a hash-based signature option.
The AI concern described in the source is less established and broader. AI systems may help researchers find mathematical techniques that change the estimated cost of solving problems used in cryptography. The source frames this as a possible challenge to assumptions behind lattice cryptography as well as older systems, not as evidence that any standard has failed. It also notes that results from AI-generated research need human verification; a withdrawn proof claim is a reminder that a plausible manuscript is not necessarily correct.
Public concern has surfaced in cryptocurrency because many blockchain transactions expose public keys, making potential key theft easier to discuss in concrete terms. On Oct. 7, Drake urged the industry to plan calmly for a possible move to what he called “bunker mode,” including using addresses whose public keys have not been exposed. The source estimates that roughly 6 million bitcoin are held in addresses with exposed public keys. That figure is an estimate in the source, not evidence that those funds have been stolen or can currently be recovered by attackers.
“Calmly begin planning for ‘bunker mode’.”
— Justin Drake, Ethereum Foundation researcher
As an affiliate, we earn on qualifying purchases.
No Cryptographic Break Is Established
The central uncertainty is whether the reported mathematical results are correct, reproducible and relevant to practical attacks. The source reports no successful compromise of RSA, elliptic-curve cryptography, ML-KEM, ML-DSA or another deployed protocol. It also does not provide technical details or independent verification of the alleged private testing by AI companies.
It is also unknown whether any newly discovered algorithm could be made practical with available computing resources, whether an organisation has found such a method and kept it secret, or how much security margin existing standards retain. Drake’s suggested window of “months not years” is a warning, not a confirmed forecast. Buterin’s comments likewise identify a concern rather than a demonstrated weakness.
As an affiliate, we earn on qualifying purchases.
Verification and Migration Reviews
The immediate next step is technical scrutiny: researchers need to verify the reported manuscripts, reproduce relevant results and determine whether any advance changes the cost of attacking real cryptographic systems. The reported withdrawal of one proof claim shows why that checking must precede conclusions about security.
Financial firms, government agencies and defence organisations will continue their post-quantum migrations while monitoring new analysis of the standards. Cryptocurrency users may also review key exposure and wallet practices, but the source does not establish that an urgent mass movement of funds is warranted. Further public evidence of a reproducible cryptographic attack—or formal assessments from standards bodies—would materially change the current picture.
quantum-resistant encryption devices
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
Has AI broken a cryptographic system?
No break is established in the source material. It reports mathematical manuscripts and warnings about possible future risks, not a successful attack on a deployed protocol.
What did OpenAI publish?
The source says OpenAI published 722 mathematical manuscripts on Oct. 6, grouped into 372 families and produced from work on roughly 4,000 problems. The claims are subject to verification.
Are NIST’s post-quantum standards confirmed to be vulnerable?
No. Buterin raised concern about lattice-based systems including ML-DSA, but the source presents no demonstrated attack against ML-DSA, ML-KEM or other NIST standards.
Should cryptocurrency users move their funds now?
Drake called for calm planning, while Buterin said he did not recommend that people scramble to move funds immediately. The source does not establish that an urgent transfer is necessary.
What evidence would clarify the risk?
Independent verification and reproduction of the mathematical results, followed by analysis showing whether they can be applied to real cryptographic protocols, would help establish whether the concern changes practical security assessments.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
