Gemini 4 Carbon 配合 AlphaProof Code 对 Signal 协议完成形式化验证
Sundar Pichai 亲自发的：DeepMind 用 Gemini 4 Carbon 给 Signal 协议生成了 600 多万行 Lean 证明，把几年的手工审计压到几周，结果还开源了。做安全或形式化验证的人应该仔细看这套智能体验证流程怎么搭的。
Google 使用 Gemini 4 Carbon 加上 AlphaProof Code，对 Signal 协议的底层代码进行数学化安全证明，验证一直下探到编译后的设备代码层，也就是真实漏洞最常藏身的环节。Gemini 4 Carbon 生成了超过 600 万行经机器校验的 Lean 证明，Lean 自动完成了绝大部分校验工作，只有约 1% 的代码需要人工复核。
原本需要安全专家数年才能完成的手工验证，被压缩到几周时间内完成。研究团队通过智能体工作流引导验证过程，覆盖了协议的全部关键代码路径，证明结论可以由 Lean 证明助手自动检查，可信度不依赖人工抽验。
Google 正在将验证结果开源给安全社区，让 Signal 协议的消息安全属性可以被独立复核。这项工作展示了 Gemini 4 Carbon 在生成大型可机检证明上的能力，与一般代码生成场景不同，Lean 的机器校验保证了产出证明的正确性，为大规模形式化验证安全软件提供了新的可行路径。
Our teams put Gemini 4 Carbon to work on a major security challenge: mathematically proving the code underlying the Signal protocol. Using AlphaProof Code, @GoogleDeepMind researchers guided agents to verify the protocol down to compiled device code, closing the gap where real bugs hide. The result: - Gemini 4 Carbon generated over 6M lines of machine-checked Lean proofs - years of expert manual verification compressed to just weeks ~1% of the codebase required human review (Lean automatically checked the rest) This is helping unlock scalable formal verification for building secure software, laying the groundwork to fundamentally shift the advantage to defenders. We’re open sourcing the results for the security community.📊 104 ⚡
This report is categorized under AITOP's "Products" desk, focusing on US frontier AI labs and Silicon Valley technology leaders. Within our cross-regional analytical framework, it directly maps to AI developer toolchains, coding agents, and engineering workflows and US-China foundation model benchmarks and commercialization comparison. Key entities and technical topics under continuous tracking include Gemini 4 Carbon, AlphaProof, Signal, Lean.
Source attribution: curated from Sundar Pichai (@sundarpichai).
