A New AI Tool From Google DeepMind Proved the Signal Protocol Is Secure
Wired

Google DeepMind used an AI framework to mathematically verify key cryptographic guarantees of the Signal protocol.
Google DeepMind researchers used an agentic AI framework named AlphaProof Code to formally verify several key cryptographic guarantees of the Signal protocol. The project, conducted as a proof-of-concept over four weeks, successfully validated portions of the cryptography that protects communications across platforms like Signal, WhatsApp, and Google Messages without finding any new vulnerabilities.
When researchers tested older versions of Signal containing known, previously fixed vulnerabilities, the AI tool correctly identified them and showed those versions could not be formally verified. The team emphasized that the analysis was limited in scope and did not cover the entire mobile app, serving instead as an early milestone in a larger development process.
While industry leaders like Signal CTO Ehren Kret praised the rigorous verification for enhancing confidence in privacy guarantees, questions remain. Experts note that it is still unclear if AI-assisted formal verification can scale broadly across the tech sector and whether it can be done at a practical cost.
Why it matters: This development demonstrates how agentic AI can accelerate software security verification, offering a potential path to mathematically prove code correctness for critical communications and financial services.



