Source time Edition date
A New AI Tool From Google DeepMind Proved the Signal Protocol Is Secure
Google DeepMind的新AI工具证明Signal协议安全
DeepMind's formal verification of the Signal Protocol gives Signal users stronger assurance, but the stated goal of verifying all the world's software shows how far such guarantees remain from everyday code.
Judgment
Signal's CTO can decide whether to extend formal verification beyond the protocol portions DeepMind checked, accepting that the four-week proof-of-concept excluded the full mobile app and came with undisclosed cost.
Basis
DeepMind research scientist Swarat Chaudhuri said the result is exciting but "there is a long distance to go" before scaling to "all the software of the world"; Signal CTO Ehren Kret said the expanded verification gives "even greater confidence" in the protocol's privacy guarantees. Mechanistically, agentic AI cut the time needed to turn software properties into formulas and proofs. The Signal project took four weeks, did not verify the entire Signal mobile app, and found no new vulnerabilities in the portions examined. On older Signal versions with known, long-fixed vulnerabilities, AlphaProof Code correctly identified them and showed the protocol could not be formally verified. Google declined to share the analysis cost and would not comment on a commercial formal-verification product.
Unclear How much it would cost to repeat the four-week Signal verification on another protocol or production codebase remains unstated.
Next Check whether the released Signal specification and proof get independent review or a follow-up verification covering the full Signal mobile app.