DeepMind AI agents formally verify Signal's code

Google DeepMind announced that its researchers used AI agents to mathematically prove key security properties of the code behind the Signal protocol. The protocol secures private messaging in apps including Signal and Google Messages.

The work used AlphaProof Code, a new multi-LLM agentic framework for formal verification. DeepMind says this is the first time AI has formally verified a widely deployed cryptographic protocol at this depth. It also says it is the first time the implementation of a widely deployed protocol like this has been formally verified by any method.

The workflow splits the labor. Researchers worked with a mix of models to turn Signal's security goals into formal specifications, and human experts reviewed every one. Agents powered by Gemini Carbon then wrote the proofs, producing more than six million lines of Lean. The agents also audited the proof and flagged about one percent of the codebase for human review. Lean's checker verified the rest automatically.

Most protocol verification stops at the design level. DeepMind says its proofs reach the compiled code running on devices. They cover properties such as keeping messages protected before and after a device is temporarily compromised. The company says the full verification took four weeks.

As a check, the team put two real, previously fixed bugs back into an isolated copy of Signal's code. In both cases the system produced a machine-checked proof that the buggy version was insecure.

The post doesn't say which parts of the protocol fall outside the proofs, whether the Lean artifacts will be published, or whether Signal took part.