You are asking the wrong question.  DMARC does not attempt to solve the
evaluator's problem with malicious impersonation.  It only solves the
sender's problem of protecting its brand image.

We know that DMARC produces false positives, and the subject earned its own
RFC.   Neither RFC7489 nor DMARCbis provide any guidance to the evaluator
for handling false positives, so he has to develop a solution on his own.
An evaluator that does not have a plan for handling false positives will be
poorly served by implementing DMARC.

We also expect that DMARC policies with p=reject cover less than 10% of all
messages received by an evaluator.  If an evaluator only blocks on Fail
with p=reject, he can assume that somewhere around 90% of all malicious
impersonation is being missed.   If he blocks on Fail with p=none, he can
reasonably assume an unacceptably high rate of false positives.  An
evaluator should consider neither outcome to be acceptable,

When DMARC successfully detects and correctly blocks a
malicious impersonation, it only blocks that one attack type.   The
evaluator is only protected until the attacker impersonates a domain
without p=reject.    An intelligent evaluator should use messages that are
blocked by DMARC (or by content filtering) to detect and block the entity
responsible for the attack.   DMARC provides no guidance for doing that
task either.

In short, your scenario is not important to DMARC.  A message with no FROM
domain has no brand threatened, so the problem is out of scope, as  Todd
correctly indicated.

A solution to the evaluator's problem will require authentication checking
on all messages, investigation of all authentication failures, and an
architecture for handling authentication failure on wanted messages.   But
this group never intended to design that solution.

Doug Foster


On Fri, Jan 31, 2025, 11:35 AM Damian Lukowski <rfc=
[email protected]> wrote:

> DMARCbis38 states
>
> DMARC operation is predicated on the input being a valid RFC5322 message
> object. For non-compliant cases, handling is outside of the scope of this
> specification.
>
> RFC5322.From can be valid and still have no author domain information.
> RFC7489 section 6.6.1 at least stated
>
> Messages with an RFC5322.From field that contains no meaningful
> domains, such as RFC 5322 [MAIL]'s "group" syntax, are typically
> ignored.
>
> So, ignore instead of fail or pass. What should DMARCbis verifiers do?
>
> _______________________________________________
> dmarc mailing list -- [email protected]
> To unsubscribe send an email to [email protected]
>
_______________________________________________
dmarc mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to