DeepMind's AlphaProof Code Formally Verifies Signal Protocol in Four Weeks
Google DeepMind announced AlphaProof Code, a multi-LLM agentic framework that formally verified key security properties of the compiled code underlying the Signal cryptographic protocol, which protects billions of private conversations in apps including Signal and Google Messages. Human experts translated Signal's security goals into Lean specifications, and Gemini Carbon-powered agents generated over six million lines of machine-checked Lean proofs in four weeks, a task DeepMind says historically takes expert teams years. Agents flagged about one percent of the codebase for human review. DeepMind says it is open-sourcing the verification artifacts.
Reference imageEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
Google DeepMind's AlphaProof Code Formally Verifies Signal Protocol in Weeks
Google DeepMind introduced AlphaProof Code, an agentic framework that formally verified Signal's cryptographic protocol within weeks. According to the post, the framework accomplished this by generating six million lines of Lean code. The announcement frames the result as a demonstration of agentic AI applied to formal verification of real-world cryptographic software. No further technical details, methodology, or independent confirmation are included in the source item.
Google DeepMind says AI framework AlphaProof Code mathematically verified Signal's protocol at code level
Google DeepMind announced on X that, for the first time, AI has mathematically proven Signal's protocol is secure down to the code level. The company said it used its AlphaProof Code framework to verify the messaging app's cryptographic system, which it says protects billions of private messages every day. The post did not provide further technical details, a paper, or independent confirmation; the claim rests on DeepMind's own statement and an accompanying link.
Google says Gemini 4 Carbon and AlphaProof Code formally verified Signal protocol code
Google teams used Gemini 4 Carbon and AlphaProof Code, with DeepMind researchers guiding agents, to mathematically prove the code underlying the Signal protocol down to compiled device code, according to the post. The work produced over 6M lines of machine-checked Lean proofs, compressing years of expert manual verification into weeks, with roughly 1% of the codebase requiring human review while Lean automatically checked the rest. The post frames the effort as unlocking scalable formal verification for secure software and shifting the advantage to defenders, and states the results are being open sourced for the security community.
Show 2 older updatesHide older updates
DeepMind's AlphaProof Code Formally Verifies Signal Protocol Cryptography
Google DeepMind announced AlphaProof Code, a multi-LLM agentic framework that combines human expertise with AI agents to formally verify software at depth. Researchers used it to mathematically prove key security properties of the actual compiled code underlying the Signal protocol, which protects billions of private conversations in apps like Signal and Google Messages. Human experts worked with models to translate Signal's security goals into formal Lean specifications, then Gemini Carbon-powered agents generated over six million lines of Lean proofs. Agents audited the proof and flagged only one percent of the codebase for human review. The verification took four weeks, a task that historically takes expert teams years. To test the system, researchers reintroduced two previously fixed bugs into an isolated version of Signal's code, and AlphaProof Code produced machine-checked proofs that the buggy versions were insecure. DeepMind says this is the first time AI has formally verified a widely deployed cryptographic protocol at this depth, and it is open-sourcing the verification artifacts.
DeepMind's AlphaProof Code Formally Verifies Signal Protocol Cryptography
Google DeepMind announced AlphaProof Code, a multi-LLM agentic framework that combines human expertise with AI agents to formally verify software at depth. Researchers used it to mathematically prove key security properties of the actual compiled code underlying the Signal cryptographic protocol, which protects billions of private conversations in apps like Signal and Google Messages. Human experts worked with models to translate Signal's security goals into formal Lean specifications, then Gemini Carbon-powered agents generated over six million lines of Lean proofs. The process took four weeks, versus years historically, and agents flagged only one percent of the codebase for human review. To test the system, researchers reintroduced two previously fixed bugs into an isolated version of Signal's code, and AlphaProof Code produced machine-checked proofs that the buggy versions were insecure. DeepMind says this is the first time AI has formally verified a widely deployed cryptographic protocol at this depth, and it is open-sourcing the verification artifacts.