What Is Zero-Knowledge Proof (ZK Proof)?

In a zero-knowledge proof there are two roles: the prover and the verifier. The prover demonstrates, for example, that they know a certain password or have performed a computation correctly, without revealing the password itself or the inputs of the computation. Such a system has three properties: true statements are accepted (completeness), it is practically impossible to produce a fake proof for false statements (soundness), and the verifier learns nothing apart from the result (zero-knowledge).

In blockchain, two main uses stand out. In privacy, networks such as Zcash prove that a transaction is valid while hiding the amount and the parties. In scalability, ZK-rollups prove to the main chain with a single short proof that many transactions were executed correctly; Ethereum nodes quickly verify this proof instead of re-running every transaction. In identity, it is also possible for a person to prove they are over 18 without revealing their date of birth. Common proof systems include zk-SNARK and zk-STARK.

Generating proofs is computationally expensive and requires expertise. Some systems require a trusted setup; if the secret data from this setup leaks, fake proofs can be produced. Programming errors in proof circuits can also lead to serious security vulnerabilities.