Perhaps Pete from Sniffer could assign a new Message Sniffer Result Code just for 
these heuristics.
We could then assign a hold based on this specific result code.

Scott Fisher
Director of IT
Farm Progress Companies

>>> [EMAIL PROTECTED] 03/19/04 03:42PM >>>
Heuristics!

This was a novel, but lame attempt at exploiting a download 
vulnerability.  This would have been 1,000 times worse if the virus 
dynamically provided a list of IP's from known infected computers.  This 
can be done, and eventually it will be done.  The kid writing Bagle has 
shown that he has some talent for coming up with new tricks, and so far 
he has come up with the best human engineering attempt, and new exploits 
for password protected files and hiding the payload outside of the 
E-mail.  It's clear to me that a person that knows this stuff has some 
experience with E-mail systems and he almost definitely works for spammers.

If he was to mix some human engineering with remotely hosted code, the 
result could be disastrous.  This attempt was lame because the exploit 
was old, long-past patched, easily detectable, and it relied on hard 
coded IP's.

Pete from Sniffer has been coding up new rules for this stuff (not all 
of his clients use Declude Virus), and if you have JunkMail Pro, it's 
easy to write a filter to block something that is IP linked to port 81.  
In the future, there will likely be little difference between what is 
necessary to block spam and viruses, and I could see when it might make 
sense to merge functionality between Declude Virus and Declude JunkMail 
to achieve a higher level of heuristics.  Full MIME parsing in JunkMail 
may very well give us many useful capabilities.  For now, I don't see 
the need as being urgent, but I've thought that such a thing as you 
described was possible for some time, and I've been wondering why it 
didn't happen.  Maybe the AV scanner companies will come out with 
command line functionality that includes content heuristics some time in 
the future.

FYI, I've found Declude JunkMail on my system tends to catch most all of 
the undetected variants that slip through in normal ZIP files early on.

Matt



Greg Little wrote:

> How will we block a virus like Bagle.Q that does not use an "auto run" 
> vulnerability?
> There's still no attachment to hand off to the mail server's virus 
> scanner(s).
> If the body was VERY standard, it could be pattern matched by Declude. 
> Add a little random action to the body (and the port used) and here we 
> go again.
>
> The latest batch of Bagle's (Q,R,S,T) can be blocked because, while 
> not a virus, it breaks the rules.
> (Auto run using a hole in MS outlook)
>
> The next version may be the same, except the user has to run it by hand.
> Just a 1 K e-mail with a link to a recently compromised PC.
>
> When will it end?????????? (or at least slow down)
>
> PS Scott,
> Thanks for the recently added Vulnerability blocking. (for Q R S & T)
>

-- 
=====================================================
MailPure custom filters for Declude JunkMail Pro.
http://www.mailpure.com/software/ 
=====================================================


---
[This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)] 

---
This E-mail came from the Declude.Virus mailing list.  To
unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
type "unsubscribe Declude.Virus".    The archives can be found
at http://www.mail-archive.com.

---
[This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)]

---
This E-mail came from the Declude.Virus mailing list.  To
unsubscribe, just send an E-mail to [EMAIL PROTECTED], and
type "unsubscribe Declude.Virus".    The archives can be found
at http://www.mail-archive.com.

Reply via email to