Google DeepMind AI proves key Signal encryption guarantees

Google DeepMind Alphabet Inc. ↘ 0.21% researchers used a new AI framework called AlphaProof Code to mathematically prove several of the Signal protocol's main cryptographic guarantees. The four-week proof-of-concept project found no new vulnerabilities in the parts of the protocol it examined, WIRED reported on November 10.wired.com

The result matters beyond the Signal app. The protocol also secures messages on Google Messages and WhatsApp Meta Platforms, Inc. ↗ 0.22%, so the analysis adds another independent check on encryption that protects a large share of the world's communications, according to WIRED. It also shows that AI agents can shorten formal verification, a long-sought but slow way of using math to show that software behaves as intended.wired.com

What the analysis found

The researchers tested AlphaProof Code on older versions of Signal that contain known vulnerabilities, all fixed long ago. The tool correctly found those flaws and showed that those versions could not be formally verified, WIRED reported. DeepMind is releasing the specification and proof it built so other verification efforts can use them.wired.com

The scope was limited. The work did not verify the full Signal mobile app, and the researchers described it as one milestone in a longer process. "We are very excited by this result, but there is a long distance to go before we are able to scale this to all the software of the world, and we want to do that," DeepMind research scientist Swarat Chaudhuri told WIRED.wired.com

The team said earlier formal verification work by Signal gave it a head start. "While prior formal verification focused on smaller portions of the code, Google DeepMind's approach enables significant expansion," Signal Chief Technology Officer Ehren Kret said in a statement, adding that the verification gives the company "even greater confidence" as AI-driven security threats grow.wired.com

A crowded field

The DeepMind work arrives while other groups are pursuing similar goals. In April 2026, Signal, the Beneficial AI Foundation and the Lean FRO launched Signal Shot, a public effort to verify the Signal protocol and its Rust implementation using the Lean proof assistant. The project aims to translate Signal's actual code into Lean and prove that it matches a verified model of the protocol, according to the foundation. DeepMind's result is separate from that project, Startup Fortune noted.beneficialaifoundation.org+2

Open questions on cost

Whether AI-automated verification can scale at a workable cost is still unclear. DeepMind did the Signal analysis for free as a research project. Google declined to say what it cost or whether it plans a commercial verification product, WIRED reported.wired.com

Chaudhuri said a key advance was a process in which AI agents contribute and correct their own work while humans review it at key points. "We don't want this sort of vibe world where AI is generating the goals and AI is generating the proofs and basically taking tests that are self-written," he said. "There's a middle ground."wired.com