Cryptographic binding

Cryptographic binding mathematically links a signature, wallet key or authentication to one document, holder or origin, not just an assertion.

Cryptographic binding means two things are linked by mathematics rather than by claim: a signature is bound to exactly one document hash, an attestation is bound to the holder's wallet key (holder binding, expressed as the cnf/confirmation claim in SD-JWT VC or as DeviceKeyInfo in mdoc), an authentication is bound to the requesting origin (origin binding in FIDO2), a payment authorisation is bound to amount and payee (dynamic linking in PSD2), and in traditional PKI a certificate binds a public key to an identity.

The pattern is always the same: whatever must not be swapped, replayed or transferred is included in what gets signed. Holder binding ties an attestation to a device key, not directly to a person; non-transferability in practice also depends on the wallet requiring the holder to unlock that key with a PIN or biometric each time. This binding is the mechanism behind sole control, non-transferable EUDI Wallet credentials and non-repudiation.