DeepMind agents lower the cost of attacking 1024-bit RSA
Original title: Google DeepMind’s AI Agents Make 1024-Bit RSA Keys Cheaper to Attack
A sandboxed rerun forged signatures in 391 core-years versus 1,380 previously; DeepMind says 2048-bit keys stay secure.
Google DeepMind, with UC San Diego and Inria, says supervised AI agents found two ways to make attacks on obsolete 1024-bit RSA cheaper. The agents reproduced an earlier signature-forgery computation in 391 CPU core-years, about 3.5 times less, without factoring the key. A second, unimplemented method would theoretically make factoring each extra 1024-bit key roughly 30 times cheaper after one costly setup. DeepMind says 2048-bit and larger keys remain secure.
Unconfirmed: The per-key factoring speedup is a theoretical estimate and has not been implemented.
- Padding adds no protection once an attacker controls about half the signed block.
- The PKCS#1 v1.5 case applies to PKCS#11 signing on devices such as HSMs.
- Forgery needs about 4 billion advance queries and does not recover the private key.
- The signature attack does not work on devices that hash messages internally.
- For a 32-byte hash, DeepMind estimates a method about 1,600 times better.
From the original article
2 min read at securityweek.com
Working with cryptographers, DeepMind’s AI agents cut the cost of forging RSA-1024 signatures and found a cheaper way to factor keys. The post Google DeepMind’s AI Agents Make 1024-Bit RSA Keys Cheaper to Attack appeared first on SecurityWeek.
By Eduard Kovacs · securityweek.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.