What is a zero-knowledge proof?

Can you convince someone you know the solution to a Sudoku without showing them a single number you filled in? That’s the idea behind zero-knowledge proofs: the other person checks that you have a valid solution, not what your solution is.

In practice, there are two sides: the one who creates the proof (the prover) and the one who verifies it (the verifier). The prover uses the secret data and the rules of the problem to produce a mathematical proof. The verifier checks it without needing access to the same secrets. “Zero knowledge” means that the proof reveals no additional information about the secret data, beyond the fact that the specific claim is true.

Why it matters

On a public blockchain, verifiability usually comes with exposed data. With the right use of zero-knowledge proofs, an app can, for example, prove that a transaction follows the rules without revealing its amount. Exactly what stays hidden depends on how the app is designed.

The technology is also linked to ZK rollups, layer 2 networks that bundle transactions and submit validity proofs to the base blockchain. That way, the base network can check a proof instead of re-executing every transaction. Validity and privacy are different things: a ZK rollup doesn’t necessarily hide your transactions.

What to watch out for

zk-SNARKs are one family of such proofs, not a synonym for the whole technology. Some systems need an initial process to generate trusted parameters, known as a trusted setup. If it’s compromised, the security of the proofs can be undermined.

Our main takeaway: a valid proof checks the rules that have been defined, not whether the entire app is secure. Coding errors, problems with crypto bridges or leaks from other data remain possible. The “zero-knowledge” label alone guarantees neither anonymity nor the safety of your money.