Google DeepMind on Nov. 10 said researchers used a new agentic framework, AlphaProof Code, to formally verify key security properties of the code behind the Signal protocol, the lab wrote. It said human experts reviewed the specifications and that agents powered by Gemini Carbon produced machine-checked Lean proofs, a process DeepMind said took four weeks. To test the system, the team reintroduced two previously fixed bugs into an isolated copy of the code, and in both cases the framework produced a proof that the altered version was insecure, DeepMind and Axios reported. Axios said the work remains a research project, not a commercial security product. Related: Gemini 4 Carbon.
The front page / AI / AI
AI / AIDeepMind Says AlphaProof Code Formally Verified Signal’s Encryption Code
DeepMind said the work took four weeks and that, when two previously fixed bugs were put back into an isolated copy of the code, the system proved those versions were insecure.
Go to the original sources
This article is an original summary written from the sources below. We encourage readers to visit them for full coverage. Spot an error? Tell us.
Get Tesla News Wire in your inbox
The Tesla news that matters to owners and investors, from software versions to delivery numbers, in one short email. Free, and you can unsubscribe anytime.
We use your email only to send the Tesla News Wire newsletter. Details in our privacy policy.