My main concern is that these spammers have gotten intelligent, many are
using AI to help them change their behavior at a pretty rapid pace, and
with less than 1 in 1 million being someone we want to see visit the
whitelist request form, we may end up needing to employ an AI just to
handle the volume of requests due to spammers, which requires the AI to
be very well tuned for decision-making and also having access to the
fleet's logs to make solid judgement calls. We're a 2 person team so I'm
trying to balance workload with high efficiency than one person is
usually expected to have.
On 2026-03-11 02:00, Hans-Martin Mosner via mailop wrote:
Am 11.03.26 um 05:31 schrieb Jarland Donnell via mailop:
...
Most of the time I get it right. On rare occasions I miss something.
It looks like I missed PolarisMail. It doesn't seem that Polaris
customers and ours converse enough that it landed on the radar, but
now they can freely. I've whitelisted their ranges.
Nobody can get automated systems to be right every time, of course.
One thing that helps a little bit is to include a whitelisting request
form URL (which is as barrier-free as possible, i.e. no account
registration, no irrelevant form fields to fill in etc.) in the SMTP
error message. We're doing that and respond very quickly to requests
(at most taking a few hours), affected senders in general seem to be
happy that their issues are resolved and that there is a personal
contact and not some black hole.
Some people can't be bothered to read SMTP error messages, those may
need to wait a while until we notice something in the logs, which may
not happen at all if they're unlucky, but happens often enough that it
still helps a bit.
Cheers,
Hans-Martin
_______________________________________________
mailop mailing list
[email protected]
https://list.mailop.org/listinfo/mailop
_______________________________________________
mailop mailing list
[email protected]
https://list.mailop.org/listinfo/mailop
