Beneficial AI Foundation
CompanyRecent news
Google DeepMind uses AI agents to formally verify Signal's secure messaging design
Google DeepMind used its AlphaProof Code multi-agent framework to formally verify Signal's post-quantum PQXDH handshake and Triple Ratchet, validating the design of libsignal's forward secrecy, authentication, post-compromise security and handshake secrecy. Gemini Carbon agents generated 6.5 million lines of Lean code, with human review of flagged sections, in response to a challenge the Beneficial AI Foundation issued in April. The proof assumes perfect cryptography and does not guarantee computational secrecy.