Google DeepMind Says AI Finds Two Major Weaknesses in 1024-bit RSA Encryption

November 7: Google DeepMind posted on X that it has published new research showing how AI agents can accelerate stress-testing of cryptographic systems. Under human guidance, its models found two major weaknesses in the traditional 1024-bit RSA encryption system. DeepMind said modern 2048-bit systems are not affected by these specific vulnerabilities, but the work shows that, as a new era of AI-driven mathematics and computation arrives, historical assumptions about computational difficulty can no longer be relied upon, and cryptographic standards must be continuously stress-tested.