Re: Verifiable Claims privacy/technology issues

On Thu, 2017-07-27 at 14:48 -0400, Tristan Hoy wrote:
> The current draft architecture for Verifiable Claims describes a
> single point of privacy failure: the identifier registry.

I'm not up to speed on the current proposals, or exchange, but it's
likely much worse than this.  

Just the nonces, and even hash of the message, in most signature schemes
already wreck any sort of privacy.  

To do this sort of thing ethically, the specification should mandate
specific named secure privacy preserving scheme that verifies the whole
certificate chain in zero-knowledge, and requires that issuers get
entirely new certificates.  

After that, you must somehow magically ask the CA to be able to attest
for the validity of a "claim" where nobody but the claim holder can even
know who the issuer is. 

Good luck with selling that!  

As I've said previous, doing "claims" correctly remains an area for
cryptographic research that is not likely to be ready for
standardization anytime soon.  I mentioned a couple recent examples of
claims done well here :
https://github.com/w3c/verifiable-claims/issues/1

Jeff