DeepMind Leverages AI Tools to Formally Verify Signal Protocol Security
1 hour ago / Read about 0 minute
Author:小编   

Researchers at Google DeepMind have employed a novel AI framework, AlphaProof Code, to complete a formal verification of the Signal protocol in just four weeks. The results indicate that no new vulnerabilities were identified in the portion of the protocol examined to date, and the framework can accurately detect issues in older versions of Signal known to contain previously identified vulnerabilities, determining that they fail verification. This achievement is regarded as a milestone in extending formal verification to various types of software worldwide. The researchers have publicly released the specifications and proofs developed for this analysis, providing support for other formal verification efforts.