I was reading the planet antispam RSS (thanks to my friend at IronPort for putting my blog on there) when I came across this article. In the article, the writer was mentioning that the image spam he was receiving contains some technical information in the subject line and some technical information in the hashbuster text in the body of the message. However, as he points out:
For the most part, this passage makes no sense and there is no connection to the subject line. But what's scary is that this is the sort of stuff that I'd normally read. And I'm pretty certain that someone, or worse, some machine knows that.
If true, this would be cause for concern indeed. Before we raise the alarm bells, we have to ask ourselves what are the odds that the user was spammed was something that he was interested in purely by chance? While I don't know what the content of spam is like, I do think that if you have a large enough sample to draw from and a spammer regularly rotates their hashbuster, then simply by statistical frequency somebody is bound to run across something they are interested in.
To do a study on this we'd need to view how many spams contain subjects and hashbusters that people are interested in, compare it to how many contain random noise and see if it has been increasing over time. For instance, how much spam do we receive that contains content that we could care less about?
If my memory is correct, if we checked our spam filters and found that 1 in 20 spam messages contained messages eerily like what we like to read, that is consistent with chance. I don't have the math handy, but it is possible to get consecutive spam messages and these results are consistent with chance. However, if we got a lot of them, then we could worry.
I got one of those the other day. It looked like it was lifted from the MSDN archives, in fact...