Google DeepMind's AlphaProof Code formally verifies Signal protocol code in four weeks
AI agents wrote over six million lines of Lean proofs, and humans had to re-check only one percent of the codebase, the company says.
What happened
Google DeepMind used AI agents to prove, with math, that the real code behind the Signal messaging protocol keeps its security promises. This is called formal verification: a mathematical proof that code does what it should. The proofs are written in Lean, a language that checks proofs automatically.
Why it matters to you
Signal's protocol protects private chats in apps like Signal and Google Messages. A proof that the actual code is correct, not just the design, lowers the chance of hidden flaws exposing many people at once.
The takeaway
AI helped prove that the code behind a major private messaging protocol is secure, in weeks instead of years.
What it means for you
The open-sourced artifacts and the specification-then-proof workflow, with Lean checking every proof, give a concrete model for verifying implementation code rather than only designs.
DeepMind says the approach could bring formal-verification-level security assurance to businesses, governments and critical infrastructure, though it says scaling needs more research.