# Introducing AlphaProof Code: a breakthrough in verifying complex software at depth

Source: [Google DeepMind News](https://deepmind.google/blog/introducing-alphaproof-code/) · 2026-11-10 · Computing
Page: https://themorning.science/story/introducing-alphaproof-code-a-breakthrough-in-verifying-complex-software-at-dept-428ac4b42c0b5f39
Kind: research · Impact: medium

## In brief

- Google DeepMind announced AlphaProof Code, a multi-LLM agentic framework for formal verification of software.
- According to the DeepMind News post, AlphaProof Code agents wrote three million lines of proofs in Lean to verify Signal's core encryption.
- The announcement was discussed on Hacker News on 2026-11-10, where it received 293 points and 110 comments.

## Why it matters

It shows that AI agents can produce machine-checked Lean proofs for a large, real-world cryptographic codebase, a possible route to verifying other critical software.

## Timeline

- 2026-11-10: Google DeepMind announces AlphaProof Code and its Lean verification of Signal's encryption code

## What science readers say

Commenters on Hacker News largely praise the achievement and see formal verification as a promising use of AI. Several raise doubts about whether AI-written Lean proofs can be trusted, whether the specification or only the implementation is verified, and whether hardware trust assumptions remain unaddressed. (sentiment: mixed)

> It’s nothing short of amazing that a huge codebase like Signal’s is apparently not only secure in a practical sense, but secure even in a strict formal sense. — p-e-w, Hacker News

> I appreciate that they reintroduced security bugs into the implementation to confirm that their process was actually catching issues. — ctrlaltdileep, Hacker News

> AI Lean is not perfect and AI can easily cheat given it is writing the lean. — r0ze-at-hn, Hacker News

## Coverage

- [Introducing AlphaProof Code: a breakthrough in verifying complex software at depth](https://deepmind.google/blog/introducing-alphaproof-code/) — Google DeepMind News, 2026-11-10

## Discussions

- [Hacker News](https://news.ycombinator.com/item?id=50439193): 293 points, 110 comments

Tags: formal verification, lean, cryptography, ai agents, signal

---

Brief generated by AI (claude-haiku-5-5) from the reporting listed under Coverage. Headlines belong to their publishers. The Morning Science – https://themorning.science