Last updated
Last updated
The idea of using zero knowledge proofs to prove claims about aspects of an individual’s identity is a major use case for zk-SNARKs.
Questions like:
“Are you over 21”?
“Do you live in the US?”
“Are you employed?”
“Are you a real human being?”
Could be answered!
Identity usecase like KYC, simply verifying the ZK proof isn’t enough!
Suppose a verifier wants to confirm that a person is over 21 using zero-knowledge technology. In theory, anyone could generate a valid ZK proof of being over 21 with a credential (containing a date of birth field) issued by any random issuer—even if the circuit verifies the issuer’s digital signature.
How can the verifier be 100% certain that the proof was actually generated using a date of birth field that comes from a verifiable credential issued specifically by the trusted issuer?
To answer this question, we need to design our in such a way that it does the following this:
Verifies if the issuer digital signature in the credential was correct?
Verifies if the credential was not revoked wrt current date?
Verifies if user/holder's digital signature is correct?
Verifies if user is above 21 (from the DOB field)?
Finally, exposes:
a) result of 4 (true/false),
b) issuer public key (or decentralised identity),
c) holder public key (or decentralised identity)
as the public signal. Now send the public signal and zk proof both to the verifier.
The verifier first checks that the exposed issuer’s public key (or DID) matches the trusted issuer’s identity, ensuring that the proof was indeed generated from a credential issued by a trusted source. Only then does it proceed to verify the ZK proof.
