It's looking more and more likely that Microsoft Security Essentials (MSE) is warning about a Trojan:DOS/Alureon.J infection when none exists.
On Sept. 16, Microsoft Answers Forum poster ElBanko said that he couldn't remove an MSE-identified infection of Trojan:DOS/Alureon.J, in spite of repeated scans and removals from multiple packages. According to ElBanko -- who appears to be quite knowledgeable:
I have run TDSS, Combofix, JunkwareRemovalTool (JRT), MS Malicious Software Removal tool, AdwCleaner, HitManPro, Malwarebytes (MBAM) and searched the registry for any related entries in a manual attempt to delete files related to it - nothing found... MSE is the only product that even identifies it but it does not remove it!
On Sept. 23, MVP Le Boule reported:
The Microsoft team that is responsible for MSE is aware of the reports of "possible" false detections of Dos/Alureon J and are working to identify and resolve any issue involving the MSE program. Anyone seeing this issue should monitor this thread.
Two weeks ago, MVP Kosh Vorlon said:
I spoke about it twice more and escalated it further yesterday with Microsoft Management - and yet again just now; however, I have no ETA yet as to when they will respond with an explanation or fix. Please be aware that they certainly know about this thread and they or I will post any updates when they become available.
Yesterday Kosh Vorlon posted:
I know it's tough, but we need to be patient. More people are posting and experiencing this and there is still no known solution or even identification of the problem by Microsoft. Also, Microsoft Support will not be able to help here. This should not be happening to this many people and they are not equipped to handle such situations. This issue needs the developers and engineers and I'm having trouble reaching the right people.
That's a lament echoed by many volunteer Microsoft MVPs who monitor the Answers Forum.
Is it a false positive or a deeper problem that MSE (and all the major antivirus scanners) can't solve? Hard to tell, but DOS/Alureon.J figured prominently in MSE circles about 18 months ago. The bleepingcomputer.com site overflows with similar complaints -- some of which are, most certainly, not false positives.
Is it possible that Microsoft re-released a bad signature for DOS/Alureon.J, one that dates back more than a year, and hasn't yet pulled the offending signature? If you're seeing alerts from MSE about DOS:Alureon.J, it would be well worth your while to join the discussion on the Microsoft Answers forum.