Certifying Zero-Knowledge Circuits with Refinement Types
Junrui Liu, Ian Kretz, Hanzhi Liu, Bryan Tan, Jonathan Wang, Yi Sun
IEEE Symposium on Security and Privacy 2024 · Day 2 · Continental Ballroom 6
This technical article delves into "Certifying Zero-Knowledge Circuits with Refinement Types," a presentation by Junrui Liu, a PhD student from UC Santa Barbara, alongside a collaborative team from institutions including UD Austin and Fraser University, and industry partners Paradise Axiom and Polychain Capital. The talk addresses a critical and often overlooked vulnerability in the burgeoning field of zero-knowledge proofs (ZKPs): the functional correctness of the underlying arithmetic circuits. As ZKPs become foundational for applications ranging from anonymous voting to privacy-preserving blockchains, ensuring that these circuits accurately implement their intended computations is paramount.
AI review
This research addresses a critical, often overlooked security gap in zero-knowledge proofs: the functional correctness of underlying arithmetic circuits. Coda, a novel DSL leveraging refinement types and machine-checked proofs, provides a robust solution. The discovery of six previously unknown, exploitable bugs in popular ZKP libraries unequivocally demonstrates its immediate and profound impact on ZKP security.