[puddintane]

Most of the time it's a brief look with a VM that the ad approval process goes thru, and honestly that may not show the malware.

A lot of the time the malware is designed to only show up in certain case scenarios such as date/time, specific version of an OS like Windows XP SP1 but not SP2. Designers come up with the most ludicrous ways of circumventing the Ad approval process, and with hundreds to thousands of new ad's per day there simply isn't an easy way to do all that testing for each single new ad being served, on top of discovering new methods used to skip the checks.

Some networks are truly terrible and just have automation systems but those networks aren't as profitable and are beginning to die out.

With deep learning networks becoming so popular I do often ponder if this type of prevention could be automated slightly better!

[basch]

safety rating of ads should exipry like edge cache