DeepMind agents lower the cost of attacking 1024-bit RSA
A sandboxed rerun forged signatures in 391 core-years versus 1,380 previously; DeepMind says 2048-bit keys stay secure.
Stories per day · 30 days
2 RSS
Relatedcryptography2post-quantum2ai1attribute-based-encryption1bilinear-diffie-hellman1bruce-schneier1bsi1classic-mceliece1clustering1defense1ecdlp1endomorphism-ring1extraction1google-deepmind1goppa-codes1hsm1isogenies1lattices1
30 stories, newest first · one card per story
Stories tagged #cryptanalysis
Adaptive clustering breaks TraintoDefend defense in model extraction
SimBreak raises signature recovery from 0% to 100% in protected networks.
Paper reassesses structural attack claims against Classic McEliece
Conditional recovery estimates fall well below targets for larger sets, but the authors say no attack is demonstrated.
Plaintext-recovery attack breaks Datta-Komargodski-Waters multi-authority ABE
The attack works in the claimed static model with one corrupt authority; a repair is proved under DBDH.
Measuring Human Cryptanalysis Effort
Matthew Green’s site estimates cumulative cryptanalyst-years spent on major public-key assumptions from published literature.
Smoothness of Shortest Isogenies to Frobenius Conjugates
Proof makes Wesolowski's subexponential attack on the supersingular endomorphism ring problem unconditional.
The Future of AI Cryptanalysis
Bruce Schneier says cryptographic over-engineering should hold up better against AI cryptanalysis than Matthew Green fears.
Provably Subexponential Algorithms for Bounded-Error LWE with Few Samples
A Kirchner–Fouque BKW variant solves bounded-error LWE subexponentially with near-linear samples.
Cryptanalysis of a Class of Ideal Secret Sharing Schemes Based on the CRT for Polynomial Rings
Cryptanalysis shows an ISIT 2026 CRT polynomial-ring secret-sharing scheme lets unauthorized subsets reconstruct the secret.
Another Historic Cipher Falls to AI
Schneier reports that AI broke another historic cipher, but the captured page lacks details.
Key-Reuse Vulnerability of Phase-Keyed Fourier-Curve Modulation: Relation Leakage and Key-Refresh Cost on Coded Links
Researchers show key reuse in phase-keyed Fourier-curve modulation leaks the key via harmonic relation moments, forcing costly refresh.
Beyond Scalar Probes: Exploiting Vector-Valued Outputs in ReLU Networks For Signature Extraction
Vector-valued Jacobian analysis improves cryptanalytic signature extraction from ReLU networks at low precision.
Cyclotomic Cosets: Hidden Subgroup and Quantum Sieving Algorithm for Prime-Power Moduli
A cyclotomic quantum sieve solves CCP and related EDCP cases quickly, but not standard LWE.
Astra and Opus just passed Turing’s other test
Cryptanalysts used OpenAI's GPT-6 Astra and Anthropic's Claude Opus 5 to crack two unsolved WWII Enigma messages.
1
Researchers Found a New Way to Break RSA that Doesn’t Require Factoring the Key
Researchers demonstrated an oracle-assisted eNFS attack that forges 1,024-bit RSA signatures without factoring the modulus.
3
Forging 1024-bit RSA signatures in nearly SNFS time
Researchers forged 1024-bit RSA signatures near SNFS speed using a signing oracle, without factoring the modulus.
There's a new way to break RSA that's faster than anything we've seen before
A new sieve-based forgery attack lowers textbook RSA security to 2^65 for 1024-bit keys.
1
Pinpointing Super-Quadratic Quantum Enumeration Speedups: Exact and Certified Evaluation of the Guessing-Moment Exponent under Product-Distribution Advice
Researchers give an exact method to measure super-quadratic quantum speedups from product side-channel advice.
Extracting CNNs in the Unknown-Architecture and Feedback-Agnostic Setting
Researchers show black-box attacks can recover a CNN's unknown architecture and its parameters.
OpenAI GPT–6 Astra breaks Enigma message that has resisted solution since 2005
OpenAI's GPT-6 Astra autonomously broke the unsolved 1941 Enigma message MVUEH, writing its own simulator, Bombe code and archive research.
2
GPT-6 Astra Solves a WWI German Radio Cipher
GPT-6 Astra solved a previously unsolved WWI German ADFGVX radio cipher, recovering a naval movement message verified against HMS Canterbury's logs.
1
The Supersingular Isogeny Problem in Time and Memory $p^{1/3+o(1)}$, Unconditionally
An unconditional algorithm solves the supersingular isogeny problem in p to the 1/3-plus-little-o time and memory.
End-to-End Hard-Label Cryptanalytic Model Extraction Using Efficient Sign Recovery
Researchers show end-to-end hard-label extraction of small ReLU MLPs, exceeding 98% label agreement on MNIST.
OpenAI's GPT-6 Astra decrypts a Nazi radio message in ten hours that went unsolved for 83 years
A Bloomberg developer used OpenAI's GPT-6 Astra to decrypt an 83-year-old Enigma-encrypted 1941 Wehrmacht radio message in about ten hours.
5