On Sun, Oct 25, 2026 at 10:59 AM John R Levine <[email protected]> wrote:
> On Sun, 25 Oct 2026, Wei Chuang wrote:
> > For interoperability, a receiver can treat an unknown algorithm signature
> > as local policy passing i.e., with a SHOULD and Authentication-Result of
> > "policy". If a vulnerability has been announced or for some operator
> > defined reason, that same local policy can set the Authentication-Result
> to
> > "softfail".
>
> See Viktor's recent messages where he points out that DKIN signatures are
> not long lived so if you think a signing scheme is broken, stop using it.
> That's straightforward and doesn't require inventing new mechanisms that
> operators are unlikely to understand. ("I sign with RSA, ED25519, and
> ML-44 using AND on all three because that makes my mail extra secure!")
>
FWIW I agree, but there still needs to be a way to express composites for
those senders who want to be extra cautious about PQC. We can look at the
TLS WG to see those arguments. IMO we should find the least complicated
way to cover their use case rather than claiming the composite proponents'
arguments are wrong.
> With your proposal, if you think a signature scheme is weak, and you add a
> new stronger one with AND, but the recipient hasn't upgraded to handle the
> strong one, the weak one passes anyway. That doesn't really help. Let's
> just leave things as they are and give people advice to sign only with
> known good algorithms.
>
I don't think the AND evaluation adds any sort of practical vulnerability
the way you are suggesting. Keep in mind the goal is to have a mandatory
set of algorithms—RSA, Ed25519, and potentially ML-DSA44—for receivers
where the latter is a placeholder for the best PQC algorithm available for
email at last call. Handling new algorithms thus remains an edge case well
into the future and will likely only affect some new PQC algorithm.
Moreover I suspect that most senders will likely use a single signature
following the earlier assumptions you and Viktor made. A threat actor has
to put in quite a bit of work to exploit the AND feature as you suggest
above: they must have a CQRC that breaks the classical algorithm and then
find a victim that doesn't recognize the PQC signature, when there are so
many other easier ways to spoof email.
-Wei
_______________________________________________
Ietf-dkim mailing list -- [email protected]
To unsubscribe send an email to [email protected]