DeepMind introduces AlphaProof Code

Google DeepMind's post on 10 November 2026 introduces AlphaProof Code, a framework that pairs people with agents to formally verify software. DeepMind says the agents proved key security properties of the code behind the Signal protocol, including that messages stay protected before and after a device is temporarily compromised, down to the compiled code that runs on devices. DeepMind says this is the first time AI has formally verified a widely deployed cryptographic protocol at this depth, and the first time an implementation of a widely deployed protocol like this has been formally verified by any method.

The post says people and a mix of models wrote the specifications, and people reviewed every specification. Agents powered by Gemini Carbon then wrote the proofs: over six million lines of Lean. The agents flagged one percent of the codebase for a second human look, and Lean checks the rest. DeepMind says the work took four weeks. To test the checker, the team put two previously fixed bugs back into an isolated copy of the code. In both cases, DeepMind says the system produced a machine-checked proof that the buggy version was insecure. DeepMind is open-sourcing the verification artifacts and links a technical paper from the post.

Who it's for: a security reader who wants the formal-verification result, not a chat product to install today.

Fits: Best for Vibe Coding stays Cursor. AlphaProof Code is a research framework for proofs, not an editor. Best for Writing stays Claude. The coding guide is unchanged. The verdict is unchanged.