API
< Back to all clusters
[TECHNOLOGY] · United States · 11 sources

started · updated

Google DeepMind uses AI to mathematically verify Signal protocol

Google DeepMind researchers have successfully used a new AI agentic framework, AlphaProof Code, to formally verify key cryptographic components of the Signal protocol. This process uses mathematical proofs to ensure software functions as intended without critical vulnerabilities, a method known as formal verification.

In a project completed in just four weeks, the AI agents generated approximately 6.5 million lines of Lean code to verify the post-quantum PQXDH handshake and the Triple Ratchet algorithm. This work validates essential security properties, including forward secrecy, authentication, and post-compromise security. Notably, the verification extended to the actual compiled machine code running on devices, rather than just the theoretical design.

The study did not uncover new vulnerabilities in the analyzed sections, but the system successfully identified known, previously patched flaws when tested against older versions of the protocol. While the scope was limited to core protocol elements rather than the entire Signal mobile application, the researchers noted the potential to scale this technology to secure other critical systems, such as medical device firmware and power grid software.

Entities

AlphaProof Code · Google DeepMind · Lean · Signal · Swarat Chaudhuri · libsignal