Anthropic's Claude Mythos Exposes Flaws in Cryptographic Algorithms

A recent breakthrough by Anthropic's AI model Claude Mythos Preview has uncovered vulnerabilities in two cryptographic algorithms, raising questions about AI's role in cybersecurity. In a significant finding, Claude Mythos was able to halve HAWK's effective key strength in just 60 hours of analysis, which would typically take years of expert human review.

Anthropic researchers used the advanced AI model to explore and experiment with cryptographic standards, identifying weaknesses in HAWK and a reduced-round version of AES. While these findings are not immediately impacting current systems, they highlight the growing role of AI in identifying potential vulnerabilities in digital security frameworks.

The discovery also underscores AI's potential to refine cryptographic research at unprecedented speeds. Anthropic plans to collaborate with academic institutions to build CryptanalysisBench, a benchmark framework for evaluating AI models in cryptography. This initiative aims to foster transparency and enable broader research into the capabilities of frontier AI models.