Anthropic's Claude Discovers Cryptographic Weaknesses in HAWK and AES, July 28

According to Anthropic, the company's Claude AI system discovered weaknesses in cryptographic algorithms today, including improved attacks against HAWK, a post-quantum encryption candidate, and a reduced version of AES, one of the world's most widely used encryption standards.

Anthropic emphasized that the discoveries do not currently threaten production systems or widely deployed encryption methods. The findings demonstrate how AI is expanding its role in cryptographic research, helping experts test stronger defenses and identify vulnerabilities before they become real-world security risks.