TLDR Tech / AI
Discovering cryptographic weaknesses with Claude
Researchers at Anthropic used Claude Mythos Preview to uncover weaknesses in cryptographic algorithms, notably improving attacks on the HAWK digital signature scheme and round-reduced AES. The AI model efficiently identified flaws previously undetected by human experts, showcasing its potential in cryptographic research. While these findings don't impact current systems, they highlight AI's role in enhancing security standards and stress-testing cryptographic designs.
Read the full story