Google DeepMind researchers introduced Alpha Proof Code, a novel agentic framework that formally verified the cryptographic protocol of Signal — a breakthrough in software security. This marks the first time AI has been used to formally verify a widely deployed cryptographic protocol at such depth. The framework combines human expertise with AI agents powered by Gemini Carbon, generating over six million lines of Lean code to mathematically prove key security properties.
The verification process took just four weeks — a task that historically takes expert teams years. The agents proved that the compiled code fulfills security properties such as message protection before and after device compromise. To test the system, researchers reintroduced two real, previously fixed bugs; in both cases, the system produced a machine-checked proof that the buggy version was insecure.
This work bridges the gap between design and implementation, where real-world vulnerabilities often hide. By automating formal verification, Alpha Proof Code enables a significant expansion of scope, offering greater confidence that Signal’s core cryptographic code delivers on its privacy guarantees.
Source: Hacker News · Summarized by HeadlinesBriefing