threat-intel
Measuring LLMs’ Ability to Perform Cryptanalysis
Medium
Summary
Researchers at Anthropic have developed CryptanalysisBench, a new benchmark to assess the ability of Large Language Models (LLMs) to perform mathematical cryptanalysis. The benchmark revealed that several LLMs, including Mythos Preview, can discover new vulnerabilities in existing cryptographic schemes and even generate novel attacks, raising concerns about the potential impact of AI-driven cryptanalysis on digital security.
Summary written automatically in our own words from the original article, which belongs to its publisher and remains the reference. It may contain errors. Sources & data