I recently had to read over some malware reports and associated logs for a confirmed malware detection and subsequent infection of a Windows asset. The logs clearly show .dll
files in a user’s AppData
folder. These .dll
files are named the same as .dll
s normally found in system32
, e.g cryptbase.dll
.
I know that in this specific instance this was definitely malware and the unpacking of the rogue .dll
s was part of the malware's normal process.
I asked about this in chat and was told that the only real credible explanation for this behaviour would be malware (as it was in this instance) or very bad programming practice, and even in that case it is a scenario that is rare.
My question is twofold; is there a scenario where .dll
files with the same name as standard system32
.dll
s be found in an user's AppData
folder for any reason other than malware or poor programming?
In addition, is it fair to treat .dll
files that are found in AppData
and appear to be copies of .dll
files in system32
, as an indicator of compromise?