AI agents from Google DeepMind formally verify parts of Signal's code
The checks cover the cryptographic library behind Signal, the encrypted messaging app, and the method has now been released as a framework.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
Google DeepMind used its AlphaProof Code system, which relies on AI agents, to mathematically verify security properties of libsignal, the core library in Signal's encrypted protocol, down to compiled machine code. DeepMind has now released the AlphaProof Code framework.
Why it matters
Reports describe the work as a landmark development for widely used cryptography and for AI itself.
Timeline
- Nov 10, 2:41 PM ET: Google DeepMind introduced AlphaProof Code, a multi-agent system that proved security properties of libsignal for one-to-one sessions.
- Nov 11, 6:54 PM ET: DeepMind released the AlphaProof Code framework and said it had analyzed Rust-based libsignal compiled into AArch64 Linux instructions.
Coverage
- DeepMind releases AlphaProof Code to verify machine code (Jae-won Choi)
- Google DeepMind uses AI to verify Signal protocol security (scworld.com)
- Gemini 4 Carbon: 6.5M Lean Proofs Verify Signal [2026] (shattered.io)
- Google DeepMind formally verifies part of Signal protocol (ua.news)
- Google DeepMind just mathematically verified key parts of Signal (Lily Hay Newman)
- Google DeepMind checks Signal encryption code with AI
- DeepMind's AI tool formally verifies parts of Signal's encryption code
More from Technology & AI
- China puts humanoid robots to work serving food and sorting parcels
- Newsom proposes sweeping AI safety rules for Washington
- Google releases Gemini 4 Neon AI model with million-token context
- Microsoft and Intel widen their cloud chip partnership
- Google DeepMind rolls out Genie 4 to select subscribers
- AI helped AMD shorten chip design work, OpenAI says