Vega: Zero-knowledge proofs for digital identity in the age of AI

Details

Vega is a zero-knowledge proof (ZKP) system from Microsoft Research that operates directly on real-world identity credentials — mobile driver's licenses, EU Digital Identity Wallet documents — allowing users to prove specific facts (age, personhood, professional status) without ever exposing the credential itself. The system generates proofs in under 100 milliseconds on commodity client devices with no trusted setup required. The technical design uses a 'fold-and-reuse' proving strategy: after the first proof is generated, repeat presentations to different services skip most of the expensive cryptographic computation. This is specifically relevant for agentic AI workflows where an AI acting on behalf of a user might need to prove credentials to multiple services in a single session. Vega works on credentials as issued — no changes required from government identity issuers. It is built in Rust and will be open sourced soon. The policy context is explicit: the EU Digital Identity framework, UK Online Safety Act age verification mandates, and similar regulations are creating demand for privacy-preserving credential checks. Current approaches require users to upload actual ID documents, creating breach risks that have already materialized in practice. For identity platform developers and enterprise security teams, Vega offers a technically mature approach to a problem that regulation is actively forcing the industry to solve. The open source release will be the real test of adoption.