🔍 Read the full analysis: The New Map For Cryptography In Finance And Defence on ThorstenMeyerAI.com
Prime made for students and young adults
- Fast, free delivery for dorm and study essentials
- Prime Video and Amazon Music included
- Member-only deals
TL;DR
OpenAI said it published 722 mathematical manuscripts on 6 October, generated with an unreleased model; the reported results include claims about faster algorithms, but they have not all been independently verified. Crypto figures Justin Drake and Vitalik Buterin have raised different concerns about AI-assisted discoveries and the assumptions behind post-quantum cryptography. No cryptographic system is reported broken, and the scale of any practical risk remains unknown.
OpenAI published 722 mathematical manuscripts on 6 October, reporting that an unreleased internal model produced them across 372 families of problems. The claims have prompted questions about whether AI could help find more efficient algorithms that challenge assumptions used in cryptography, but no cryptographic system has been shown to be broken, and the reported mathematical results remain subject to checking.
The source account says the model worked on roughly 4,000 problems, with about three hours of ChatGPT Pro compute per result on average. The manuscripts include claimed results concerning the Unique Games Conjecture, Hilbert’s tenth problem over the rationals and a zero-free region for the Riemann zeta function. Those claims should not be treated as settled: verification is ongoing, and the account notes that OpenAI withdrew a claimed proof related to the Hodge conjecture after a reported sign error.
For cryptography, the concern is not simply that AI can produce mathematical writing. Some reported results concern computational complexity: the source describes claims involving faster integer multiplication and Fourier transforms, and an algorithm for 3SUM with a running time of about n^1.9992. The latter result was in a paper by Virginia Vassilevska Williams and Josh Alman; the account says an Anthropic model supplied the key idea. These examples do not establish that practical encryption can be defeated.
Computer scientist Scott Aaronson observed that cryptography was conspicuously absent from the 722 manuscripts. According to the source, people Aaronson consulted said AI companies had begun discreetly testing whether their internal models could break important protocols. That is a reported account, not a public demonstration of a successful attack. The distinction matters: testing a system, proposing a mathematical approach and recovering a real key are different milestones.
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 Cryptographic Assumptions Matter
Financial networks, intelligence services and defence systems rely on cryptographic tools to protect communications, authenticate users and secure transactions. If a new algorithm made a widely used mathematical problem substantially easier, systems depending on that problem could require changes. The potential concern extends beyond cryptocurrencies, whose public transaction records can make some exposures easier to observe.
The discussion also complicates planning for the quantum threat. A sufficiently capable quantum computer running Shor’s algorithm could break RSA and elliptic-curve cryptography. That risk has led governments and companies to prepare replacements. An AI-assisted mathematical advance would be different: it could run on conventional computers and might not be visible before its discovery or use. This is a possibility raised in the source material, not evidence that such an algorithm exists.
For financial institutions and defence agencies, the practical implication is to avoid treating any migration plan as proof that future systems are invulnerable. Cryptographic choices rest on assumptions about how difficult certain mathematical problems are. Reviewing those assumptions and planning for updates can matter even without a demonstrated break, while claims of immediate compromise would go beyond the available evidence.
quantum-resistant cryptography hardware
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
From Quantum Migration to AI
Governments and companies have been preparing for quantum computers because Shor’s algorithm threatens public-key systems such as RSA and elliptic-curve cryptography. In August 2024, the US National Institute of Standards and Technology standardised ML-KEM for establishing encryption keys, ML-DSA for digital signatures, and SLH-DSA, a signature scheme based on hash functions. These standards address the anticipated quantum threat; their adoption is not a guarantee against every possible future mathematical advance.
The source argues that lattice-based systems, including ML-KEM and ML-DSA, could merit scrutiny if new mathematics changes the difficulty of problems they rely on. It does not report that these standards have been broken or that their security has been invalidated. Hash-based cryptography has different foundations, but the source likewise does not establish that any family of cryptographic systems is safe against all AI-assisted discoveries.
Blockchain security has made the debate more visible. On 7 October, Ethereum Foundation researcher Justin Drake urged the industry to plan calmly for “bunker mode,” including moving funds to addresses whose public keys have not been exposed. The source says roughly six million bitcoin sit in addresses with exposed public keys, but does not supply a measurement method or a comparison period for that estimate. On the following day, Ethereum co-founder Vitalik Buterin cautioned against an immediate rush to move funds and highlighted possible risks to lattice-based cryptography.
““Calmly begin planning for ‘bunker mode’.””
— Justin Drake, Ethereum Foundation researcher
What the Evidence Has Not Shown
No demonstrated cryptographic break is described in the source material. It does not identify a working AI-generated method for recovering private keys, defeating a deployed protocol or breaking NIST’s post-quantum standards. The manuscripts’ mathematical claims are also not all verified; the reported withdrawal over a sign error illustrates why review is needed.
It remains unclear whether AI systems can find practical cryptographic attacks, how often they may do so, or whether any organization has discovered and kept such a result secret. The source’s account of discreet testing relies on unnamed sources. It also offers no independently confirmed evidence that a powerful method is already being used against financial, intelligence or defence systems. The scale and timing of any potential threat are unknown.
Verification and Security Reviews
The immediate next step is mathematical and cryptographic scrutiny: independent researchers need to check the manuscripts and determine whether any claimed algorithm changes the practical security of real systems. A faster solution to a theoretical problem would not automatically translate into an attack; implementation details, required computing resources and the size of the improvement all matter.
Organizations responsible for finance, intelligence and defence can continue their post-quantum migration while monitoring verified research and reviewing how their systems could be updated if assumptions change. The source provides no timetable for a confirmed AI-driven cryptographic advance, and no official new standard or migration deadline is reported. For now, the development is a reason for careful evaluation, not evidence that users should abandon existing protections.
Key Questions
Did AI break encryption?
No reported break is confirmed. The source describes AI-generated mathematical manuscripts and reported testing of protocols, but no demonstrated recovery of a cryptographic key or defeat of a deployed system.
What did OpenAI publish?
OpenAI published 722 mathematical manuscripts on 6 October, grouped into 372 families. The source says an unreleased internal model generated them, but the claims require independent checking.
Are post-quantum standards compromised?
The source does not show that NIST’s standards have been broken. Vitalik Buterin raised concerns about lattice-based systems, including ML-DSA, but that is a warning about possible mathematical assumptions, not proof of a vulnerability.
Why are cryptocurrency users part of the discussion?
Blockchain activity can expose public keys and make funds associated with particular addresses visible. Justin Drake recommended planning for addresses with unexposed keys; Buterin said users should not rush to move funds today.
What should institutions watch for next?
They should watch for independently verified results showing a practical improvement against cryptographic problems used in real systems, including details about required computing resources. No such result or official new deadline is reported here.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
