🔍 Read the full analysis: What AI Mathematics Means For Cryptography In Finance And Defence 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
OpenAI published 722 mathematical manuscripts on Oct. 6, while a separate AI-assisted result challenged a long-standing algorithmic conjecture. The developments have prompted crypto figures to discuss risks beyond quantum computing, but no cryptographic protocol has been publicly shown to be broken, and the implications for post-quantum standards remain uncertain.
AI-produced mathematical work has prompted crypto researchers and industry figures to question whether advances in algorithm discovery could challenge assumptions behind widely used security systems, including emerging post-quantum methods. The reports describe no cryptographic protocol being broken; the concern is that a better algorithm could weaken a system without the visible countdown associated with quantum hardware.
On Oct. 6, OpenAI published 722 mathematical manuscripts across 372 families, generated by an unreleased internal model from roughly 4,000 problems, according to the source material. It says the work included claims involving the Unique Games Conjecture, Hilbert’s tenth problem over the rationals and a zero-free region for the Riemann zeta function. Those claims require checking; one reported proof, concerning the Hodge conjecture for products of K3 surfaces, was withdrawn after a sign error was identified.
For cryptography, the source highlights algorithmic results rather than the famous conjectures. Computer scientists Virginia Vassilevska Williams and Josh Alman reported a result for 3SUM with running time of about n^1.9992, challenging a long-held belief that the problem required roughly quadratic time. The source says an Anthropic model supplied the key idea. It also cites Scott Aaronson’s account of claimed improvements to integer multiplication and the Fourier transform. These results do not by themselves show that encryption or digital signatures can be defeated.
Aaronson also noted that cryptography was absent from OpenAI’s published collection. He said his sources indicated that AI companies were discreetly testing whether internal models could break important protocols. That account is not a public demonstration of a successful attack, and the source provides no protocol, test results or independent verification. The distinction matters: finding faster algorithms can alter security estimates, but cryptographic impact depends on whether a method works against a deployed system at practical scale.
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.
Why Finance and Defence Should Track It
Financial institutions, intelligence agencies and militaries depend on cryptography to protect transactions, communications, stored data and authenticated instructions. Their migration planning has largely treated quantum computing as the central long-term threat to public-key cryptography. AI-assisted mathematics introduces a different uncertainty: a mathematical weakness might be found on ordinary computers, and a discovery could remain private rather than appearing in public hardware milestones.
That possibility does not establish that current systems are unsafe. It does underline the need to distinguish standardised algorithms from proven security: cryptographic schemes are supported by extensive analysis, but their hardness assumptions are not generally mathematical proofs that no efficient attack exists. For banks and defence organisations, the practical issue is whether risk assessments and migration plans account for algorithmic breakthroughs as well as quantum progress, without reacting to unverified claims as if a breach had occurred.
The concern also bears on post-quantum cryptography. NIST’s lattice-based standards are designed to resist known quantum attacks, but the source raises the question of whether future mathematical techniques could affect assumptions behind lattices. It presents this as a possibility, not evidence that ML-KEM or ML-DSA has been compromised.
As an affiliate, we earn on qualifying purchases.
Quantum Migration Meets a New Unknown
Quantum risk has a comparatively clear mechanism: a sufficiently capable, error-corrected quantum computer running Shor’s algorithm could break RSA and elliptic-curve public-key cryptography. In August 2024, NIST standardised post-quantum replacements including ML-KEM for key establishment, ML-DSA for digital signatures and hash-based SLH-DSA. Organisations have been preparing to migrate before large-scale quantum attacks become feasible.
The source contrasts that hardware-based warning with AI’s potential role in discovering algorithms. Quantum progress can be followed through hardware and engineering milestones; a mathematical result might be developed privately, with no public sign of the work. The source also records two different reactions in the cryptocurrency sector. On Oct. 7, Ethereum Foundation researcher Justin Drake urged planning for “bunker mode,” while Ethereum co-founder Vitalik Buterin said he did not advise people to rush to move funds. Buterin identified lattice-based systems, including ML-DSA, as an area of possible risk rather than a system known to be broken.
No Public Cryptographic Break Reported
The source reports no confirmed break of RSA, elliptic-curve cryptography, ML-KEM, ML-DSA or another deployed protocol. It does not provide details or independent verification of the reported private testing by AI companies, nor evidence that any model has produced an attack that works in practice. The extent to which the cited mathematical results can be checked, reproduced and applied to cryptography is also unsettled.
Drake’s estimate about a possible ECDSA break is explicitly a warning about a worst-case scenario, not a measured capability. The source’s figure of roughly six million bitcoin in addresses with exposed public keys is a reported estimate; it does not mean those holdings are currently vulnerable to a demonstrated AI attack. The year attached to the October dates is not specified in the supplied material.
Verification and Security Reviews Ahead
The immediate next step is independent mathematical review of the AI-produced manuscripts and algorithmic claims, followed by careful testing of any result relevant to cryptographic workloads. A claimed improvement in a general mathematical problem would need to be translated into a practical method against a specified cryptographic scheme before security assessments could change.
Financial and defence organisations are likely to continue quantum-readiness work while monitoring research into AI-assisted algorithms. The source gives no timetable for public results from the reported private tests, and no agency or standards body is described as changing its guidance. For now, the development is a reason to scrutinise assumptions and evidence—not a notice that existing encryption has failed.
Key Questions
Has AI broken a cryptographic system?
No confirmed break is reported in the source material. It describes warnings and possible private testing, but no demonstrated attack against a named deployed protocol.
What did OpenAI publish?
OpenAI published 722 mathematical manuscripts in 372 families on Oct. 6, according to the source. The manuscripts contain claims that require expert checking, and at least one reported proof was withdrawn after an error was found.
Could AI threaten post-quantum cryptography?
It is a possibility raised by the source, particularly for schemes based on lattices, but there is no reported evidence that NIST’s post-quantum standards have been broken. Their practical security would depend on a specific, validated attack.
Should cryptocurrency holders move their funds?
The source quotes Justin Drake urging preparation and Vitalik Buterin advising against scrambling to move funds immediately. It reports no confirmed attack, so the comments are different risk assessments, not proof that a transfer is necessary.
Source: ThorstenMeyerAI.com
Columbus Day / Indigenous Peoples' Day Picks
long weekend sales
As an affiliate, we earn on qualifying purchases.
