Aug 06 2020 10:12 AM
Aug 06 2020 10:12 AM
In the Action Center I would like to automatically reject the pending actions "Block URL" and "Soft delete emails". I know this is not natively supported, but is there way to automate this using PowerShell, Power Automate, Security Graph API or something else?
Aug 09 2020 03:07 AM
Aug 11 2020 12:05 AM - edited Aug 11 2020 02:16 AM
Hi @Evald Markinzon
We want to automate as much as possible, so for these actions we want ZAP and native SafeLinks to automatically handle deleting emails and blocking URL.
These extra pending actions are being generated in large volumes and create too many false positives, so we've decided to always reject them and perform these actions manually when necessary.
To avoid alert fatigue and to better highlight the other Actions we need to evaluate, we want to automatically reject these actions so they don't flood the list.
Aug 13 2020 12:34 AM
@Joachim83 - Zap does automatically act on emails and is the source of many of these alerts. Zap will perform the action chosen for that threat in the appropriate policy - so if you have phish set to junk, zap will auto-junk it.
However, there are cases where Zap may not remediate the email - e.g. emails that were over 48 hours old with the malicious url/file, phish emails where the user/organization had 'override policies' (e.g. safe sender, safe domain, ETRs, etc.), plus emails from similar emails but with different malicious links/files that don't get identified. I definitely hear your request for clearer needed actions. For the moment want to make sure admins review and give us feedback when they disagree with aspects of the investigation. As Evald said - if you simply ignore the investigations they'll expire.
Safe links does auto-block the links today if you're applying the policy to your usrs - the action from the investigation is redundant right now.
I'd suggest you definitely review any 'User compromise' and 'URL verdict change' investigation at a minimum. These are high severity because they are situations where the user may be compromised - so the other details in the investigation are particularly worth reviewing (user evidence, bad URLs, etc.).
Aug 13 2020 12:47 AM
We have a Power Automate job that automatically trigger a full MDATP antivirus scan and investigations on any computer a user has logged in to the last 3 days when a High Threat Management Alert has been triggered by a user. These we do check after they have completed.
However the URL block actions are redundant like you mention as SafeLinks have already blocked them, so we want them automatically rejected.
And if ZAP doesn't automatically delete e-mails we will create manual remediation jobs, so we want to automatically decline the delete email actions as well.
Aug 17 2020 02:03 AM
Beyond what was already discussed regrading the redundant actions, and your preference for manual investigation. Could you please elaborate more on the false-positives issue that you mentioned?
Would be great if you could provide any concrete details so we can investigate this further and improve our investigation logic accordingly.
Aug 17 2020 02:21 AM
In generel e-mail identified as Phishing starts an investigation that creates 4 different E-mail Soft Delete actions and one Block URL action for every URL in the email.
Some of these e-mails are false positives, as an example e-mails from the shipping company postnord.no is regularly detected as phishing. It is very easy for someone to approve these pending actions withouth proper investigation and then legitime e-mails and URL's are deleted and blocked for the entire organization.
If we just let them be and await for them to expire, then the other actions we actually want to look into are harder to notice in the sea of pending actions.