On Wed, Jun 24, 2020 at 3:08 PM Kathleen Wilson via dev-security-policy <
dev-security-policy@lists.mozilla.org> wrote:

> I have updated the following section of the wiki page to incorporate
> feedback that I received from representatives of ACAB'c.
>
>
> https://wiki.mozilla.org/CA/Audit_Statements#Verifying_ETSI_Auditor_Qualifications
>
> I will greatly appreciate it if those of you familiar with ETSI audits
> will review it and provide feedback.


While browsers have certainly discussed requiring ACAB’c membership, and
rejecting audits from CAs that are not members, I haven’t seen much
transparency yet from ACAB’c about how they supervise and/or review audits
in the same way WebTrust does.

It also runs a risk with respect to trusting a third-party organization to
validate the qualifications. I could totally understand and agree that if
ACAB’c membership was mandatory, because of some clear value to browsers
such as Mozilla, this would make sense, but right now it seems like it may
be premature?

For example, as noted by ACAB’c itself, accreditation is with respect to
ISO 17065 and eIDAS Art3.18, but that provides zero guarantees with respect
to certificates, the BRs, or to the ETSI EN 319 403 provisions. For
example, if a notified scheme under eIDAS made use or certificates in a way
that is directly in conflict with Mozilla, those auditors could still be
seen as skilled to assess Mozilla’s requirements. That seems... odd?

I would suggest that, for the time being, ACAB’c isn’t a shortcut. I
realize that means more work for Mozilla, and broadly for the industry, but
it might provide an opportunity for ACAB’c to focus on whether the goal is
to support eIDAS audit schemes and accreditation, or whether it is to
provide browsers equivalent confidence and focused collaboration in the way
the WebTrust TF had engaged in. That isn’t to suggest the auditors might
not also provide eIDAS audits, but it seems a real missed opportunity for
auditors to more proactively engage and ensure needs like Mozilla’s are met.

I realize the template is a valuable step, but I don’t think ACAB’c
membership alone is equivalent to the assurances browsers get from WebTrust
licensure, and I worry that the “simple” step will encourage auditors that
are not appropriately qualified for browser use cases.

I know this means considerably more work for Mozilla, to disregard ACAB’c
for the time being, and so I don’t want to seem dismissive of that. If
anything, my hope is it might encourage more engagement by ACAB’c here, and
a more careful evaluation of membership, criteria, and focus, such that the
end state is we can skip the NAB/CAB validation and instead be confident
that ACAB’c membership is a useful and positive signal of auditor
qualification. I just don’t know that we’re there yet, and I worry ACAB’c
might be just a little too eager, and a little too generalized, right now :)

>
_______________________________________________
dev-security-policy mailing list
dev-security-policy@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-security-policy

Reply via email to