🔍 Read the full analysis: The Changing Cryptography Behind Finance And Defence on ThorstenMeyerAI.com
Get the little things that make your day 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
An OpenAI release of 722 AI-produced mathematical manuscripts and reports of faster algorithms have prompted renewed questions about the assumptions behind cryptography. Ethereum researchers have urged preparation and caution, but the source reports no demonstrated break of a cryptographic system. The potential impact on finance and defence remains uncertain, especially because a useful algorithm could be kept secret.
OpenAI published the manuscripts on 6 October, grouped into 372 families and generated by an unreleased internal model working on roughly 4,000 problems. The source 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. These are research claims that require checking, not settled mathematical results.
For cryptographers, the more relevant developments described in the source concern computational speed. It points to work on integer multiplication and Fourier transforms below the n log n threshold, as well as a roughly n^1.9992 algorithm for 3SUM. The 3SUM result appeared in a paper by Virginia Vassilevska Williams and Josh Alman; the source says an Anthropic model contributed the key idea. These results do not themselves break encryption, but they illustrate why researchers are watching for algorithms that could weaken assumptions about how hard certain computations are.
The source also reports that AI companies have begun discreetly testing whether internal models can break cryptographic protocols, citing computer scientist Scott Aaronson. It says a claimed OpenAI proof about the Hodge conjecture for products of K3 surfaces was withdrawn after a sign error was identified. That episode underscores the difference between generating a plausible result and independently verifying it.
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 Cryptographic Migration Plans
Governments and companies have been preparing for the possibility that sufficiently capable quantum computers could use Shor’s algorithm to break RSA and elliptic-curve cryptography. The concern raised by AI-assisted mathematics is different: a new algorithm might run on conventional computers and could target assumptions behind more than one family of cryptography. Unlike progress in quantum hardware, which can be tracked through public engineering milestones, an algorithm might be discovered and withheld.
That possibility matters to banks, intelligence agencies and defence organisations, whose communications, identity systems and digital signatures depend on cryptographic protections. It does not establish that these systems are currently vulnerable. It does raise a planning problem: migration to a replacement standard may not settle every risk if the replacement relies on mathematical problems that later prove easier than expected.
For digital assets, the exposure can be unusually visible because public keys and transactions are recorded on open networks. The source cites a report estimating that about 6 million bitcoin are held in addresses with exposed public keys. That figure is a reported count; the source gives no date or comparison baseline for it, and it is not evidence that those funds can now be recovered by an attacker.
As an affiliate, we earn on qualifying purchases.
Quantum Standards And New Warnings
The established migration effort centres on standards published by the U.S. National Institute of Standards and Technology in August 2024. They include ML-KEM for establishing encryption keys, ML-DSA for digital signatures and SLH-DSA, a signature scheme based on hash functions. These are intended to address the quantum threat; the source questions whether confidence in lattice-based methods should be treated as permanent, not whether the standards have been broken.
On 7 October, Ethereum Foundation researcher Justin Drake urged the industry to plan for a possible “bunker mode,” recommending that users move funds to addresses whose public keys have not been exposed. He warned that, in a worst-case scenario, ECDSA might be broken before the arrival of a large quantum computer. Ethereum co-founder Vitalik Buterin cautioned against a rush to move funds, while arguing that lattice-based methods and related technologies also merit scrutiny.
The two positions are warnings about risk, not reports of a working attack. Buterin’s historical point is that mathematical techniques have sometimes made problems once assumed to be costly substantially faster to solve. Whether comparable advances exist for elliptic curves or lattice problems remains unknown.
““IMO it is now reasonable to brace for the possibility that ECDSA breaks before qday, in the worst case in months not years.””
— Justin Drake, Ethereum Foundation researcher
No Public Cryptographic Break Reported
The source describes no verified attack on RSA, ECDSA, ML-KEM, ML-DSA or another deployed cryptographic protocol. It also does not provide the full manuscripts, independent reviews of the relevant algorithmic claims, or evidence that an AI system has recovered a private key from a real system.
It is unclear whether the reported AI-company tests have found useful cryptographic attacks, how their results were validated, or whether any discoveries have been kept confidential. The source’s account of the 722 manuscripts also gives no year for the October dates. The extent to which AI methods could improve algorithms against specific cryptographic problems remains an open research question.
Verification And Migration Decisions
The immediate next step is independent mathematical review of the AI-generated work and the reported algorithmic advances. Researchers and standards bodies would need to determine whether any result changes the estimated security of a specific cryptographic scheme before institutions alter their systems on that basis.
Financial institutions and government agencies are expected to continue evaluating post-quantum migration plans, including the standards already published by NIST. The source gives no announced deadline or policy change tied to the AI results. For cryptocurrency users, Drake and Buterin’s comments frame a debate over preparedness rather than an instruction supported by a confirmed break; further technical evidence and guidance from the relevant networks would be needed to establish what action is warranted.
Key Questions
Has AI broken encryption used by banks or governments?
The source reports no verified break of a deployed cryptographic system. It describes new mathematical work and warnings about possible future algorithmic advances.
What did OpenAI publish?
On 6 October, OpenAI published 722 mathematical manuscripts in 372 families, generated by an unreleased internal model. Their claims require independent checking, and the source says one claimed proof was withdrawn after a sign error was found.
A sufficiently capable quantum computer could use Shor’s algorithm against RSA and elliptic-curve systems. The AI-related concern is that improved mathematics might produce faster algorithms on ordinary computers; no such cryptographic break is established in the source.
Should cryptocurrency holders move their funds now?
The source records differing cautions: Justin Drake recommended planning for a protective “bunker mode,” while Vitalik Buterin said he did not recommend scrambling to move funds immediately. It reports no confirmed attack or universal direction for holders.
Are post-quantum cryptography standards also under discussion?
Yes. Buterin specifically raised questions about lattice-based methods, including ML-DSA. The source presents this as a risk to examine, not evidence that NIST’s standards have been broken.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
