Re: When is "phone home" ok, if ever?

On Thu, May 8, 2025 at 1:34 PM Kim Hamilton <kimdhamilton@gmail.com> wrote:
> Joining back from some side discussions, I think it's clear that "phone home" needs a clearer definition.  We all agree that the issuer is authoritative for the status of credentials they issue — the difference is on how status checks occur, and what sort of tracking it enables.

Yes, I think the above is one of the things we need to do. It's
somewhat analogous to the  question of "Exactly what does
'decentralized' mean, and why do we keep arguing about it?" we had to
answer several years ago with the DID Rubric. Turns out there are many
dimensions of "decentralization" and the word was far more nuanced
than we thought it was at first. "Phone home" seems to have the same
problem.

The discussion so far has been excellent; deeply insightful, which is
no surprise given the folks involved in this community. I'm going to
try to go back and engage directly with each email, but here are the
themes I'm picking up based on what many people have said so far:

* The "digital wallet" (or the app acting on behalf of the individual)
MUST serve that individual first and foremost. This is similar to the
trust we have in browsers today to not expose our information to bad
actors. It breaks down a bit with browsers, though, as some of them
passively aid in massive data collection in order to provide us with a
"free" Web. So, we want better assurances with wallets because they
really do deal with highly personal and sensitive information.

* "Phone home" has multiple interpretations and nuances. There is
acceptable phone home, such as when a verifier retrieves a Bitstring
Status List in a way that provides good anonymity for a particular
holder in the list. There is also unacceptable phone home, such as a
verifier contacting the issuer and reporting in on specific holder
behaviour that was never consented to by the holder.

* Real-time tracking is being asserted as strongly out of scope for
VCs and wallets.

* Pingbacks for an individual acting in an official government
capacity, which might require auditing, is debatably ok if done by a
verifier that is configured to do so. There are good arguments for and
against, and I hope to explore this particular point in a future
email... but the core point seems to be: auditing is a separate
process that happens after verification, and we should keep it layered
in that way.

* These systems need to work offline in emergency situations, so
building anything that requires a network connection to provide base
functionality is asking for trouble (a broken system).

* During verification, there is only "good enough" freshness for a
particular use case. The "good enough" timeframe can be different per
use case.

Those are all great insights, and I am still going through Carsten's
excellent write up on the different expectations between credentials
issued to individuals vs. credentials issued to employees.

We need to capture all of this -- it's important guidance to others
building credentials and wallets for different ecosystems.

-- manu

-- 
Manu Sporny - https://www.linkedin.com/in/manusporny/
Founder/CEO - Digital Bazaar, Inc.
https://www.digitalbazaar.com/

Received on Saturday, 10 May 2025 21:00:19 UTC