M365 - Sudden Increase in outgoing "phishing" false positives

oikjn

Ars Scholae Palatinae
969
Subscriptor++
Any M365 admins out there seeing a sudden surge in outgoing mail getting flagged as phishing/spam starting a few hours ago in the OUTBOUND direction?

I had the outbound filter rule set to send me a copy of all flagged emails and in the last couple of YEARS I think maybe I've seen one or two emails come in. Today I'm at 28 so far. I"ve checked most of them and they are just normal email exchanges between employees and outside mailboxes.

In the Defender realtime detection screen, it shows as "Latest Threats" = "Phish / Normal, Spam" and the "Detection Technologies" = "General filter, Mixed analysis detection".

The Sender IP is our local IP. I used mxtoolbox.com to check blacklists for our IP and our domain and nothing has changed on that end nor in our DNS records. So I don't think its me, but...
 

oikjn

Ars Scholae Palatinae
969
Subscriptor++
Haven't noticed any. Are you using a hybrid deployment or just 365?
just 365.

It seems to have chilled back out again. It seems like it was just a 12 hour spurt and my rule didn't block the traffic and just forwarded me a copy (assuming it wasn't also incorrectly flagged on the inbound side if that end recipient was also using M365).
 

mrkag

Smack-Fu Master, in training
1
I believe it's related to the Microsoft Tech blog post titled "Announcing New DMARC Policy Handling Defaults for Enhanced Email Security". I have been seeing DMARC rejects in both our commercial and GCC tenants, starting about a month ago.

It's irritating that some email admins don't understand how DMARC and SPF work, or even what spoofing means.
 

oikjn

Ars Scholae Palatinae
969
Subscriptor++
I think what @moosemaimer linked was correct and my experience was pretty much exact to what was described in that link.

@mrkag while I agree that there are far too many "admins" out there that don't understand DMARC let alone SPF or DKIM, in this case, I can say it wasn't the cause here. This was all internal authenticated users sending emails from my domain which is setup properly with DMARC/SPF/DKIM and the quarantine report confirmed they were all passed... that being said, they were all authenticated users sending emails using Outlook, so I wouldn't put it past M$ to have flagged their own servers as spam sources for whatever reason.