[Google DeepMind Claims AI Discovers Two Major Weaknesses in 1024-bit RSA Encryption] Google DeepMind posted on its X platform that it has released new research demonstrating how AI agents can accelerate stress testing of encryption systems. Under human guidance, its model discovered two major weaknesses in the traditional 1024-bit RSA encryption system. DeepMind stated that modern 2048-bit systems are unaffected by these specific vulnerabilities, but this work shows that with the arrival of a new era of AI-driven mathematics and computation, it is no longer possible to rely on historical assumptions about computational difficulty; continuous stress testing of cryptographic standards is essential.