Anthropic's Claude AI Discovers Flaws in HAWK and AES Encryption Algorithms

Anthropic reported that its Claude artificial intelligence model has identified new mathematical vulnerabilities in cryptographic algorithms used to secure digital systems, prompting fresh discussions surrounding the long-term future of encryption standards.
The research demonstrated improved attack vectors against HAWK as well as variants of the Advanced Encryption Standard (AES). Anthropic noted that no live production systems are currently impacted or at immediate risk from the discovered weaknesses.