[35421] in bugtraq
Re: Is predictable spam filtering a vulnerability?
daemon@ATHENA.MIT.EDU (Chris Brown)
Tue Jun 22 02:53:15 2004
Message-ID: <40D6E158.7030706@wavetex.com>
Date: Mon, 21 Jun 2004 08:23:36 -0500
From: Chris Brown <chris@wavetex.com>
MIME-Version: 1.0
To: bugtraq@securityfocus.com
In-Reply-To: <009601c45476$00737fe0$650aa8c0@aaronxp>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Aaron Cake wrote:
>Imagine if I decided to use a spam fitler against someone else...I make an
>email that contains known rejected words. I send that email, setting the
>"FROM" address and header to be that of my victim. If I send out hundreds of
>these messages, I can use someone else's spam filter to mail-bomb my victim
>with "rejected" messages.
>
>
This definitely is a problem, but not just like you describe. Spammers
themselves do the same thing based simply on the way SMTP works. At the
ISP I work for we've had tons of problems with a spammer (or multiple
spammers) using an address from one of our domains as the sender and
then a bad recipient address and sending thousands of messages. The
recipient server bounces messages back to the apparent sender,
effectively mail-bombing them like you describe. Its even worse when
the victim address doesn't actually exist because then the postmaster
address at both ends gets all kinds of warnings as well.
I've seen some ISPs have started to block mail sent by <> to put an end
to this sort of thing, but that doesn't seem like a very good solution
as it probably blocks legitimate bounce messages.
Chris
