Anthropic's Claude model uncovers long-standing bugs in two encryption algorithms
On Tuesday, Anthropic announced that its Claude Mythos Preview AI discovered mathematical vulnerabilities in a pair of cryptographic algorithms that had gone unnoticed by experts for years. According to the firm, the first exploit notably compromises HAWK, a digital signature system designed for post-quantum security. The second finding reveals a novel method for attacking round-reduced AES, which is the world's most common symmetric cipher.
by shortkt.com✓