We recently came across an InfoWorld.com item suggesting that an HTML 'bug' implanted in spam could be a major boon to malicious hackers. The technique mentioned involves embedding a link to a tiny, one-pixel image on the spammer's server.
When victim retrieves the message, his e-mail client automatically fetches the image off the spammer's server in order to display it in the message window. Since the image is miniscule, the victim never sees it and never suspects that his client is communicating with a remote server.
The logs then tell the spammer which e-mail addys or IPs connected to his server looking for the image, which in turn tells him which e-mail addresses are valid, and thus he keeps them on his victim list.
All this is pretty smarmy; it means that spammers can verify valid e-mail addresses with decent accuracy. The normal defensive techniques of not following any link in a spam message and never replying to the decoy "remove me" address (which, far from getting you removed, only confirms that your address works), would no longer be effective.
But we're not convinced that the specific technique cited - merely embedding a one-pixel image on the body of a spam message - could give malicious hackers remote access to a network or machine, or anything else that they couldn't get more easily through other means.
However, the possibility is suggested in the InfoWorld piece: "It's just a matter of time [before] someone [can] figure out how to use these things against people or corporations," the paper quotes Sharon Ward, director of enterprise business applications at Hurwitz Group, as saying.
But what can be done, we wonder? Certainly one gets a list of valid IPs out of it; but doing it that way would be a chore, whereas a normal port-scan according to IP ranges can be automated and run completely in the background.
Such HTML bugs would also be useful to garner a list of valid IPs in a particular domain; but there again, tools exist which require less effort to do the same (a reverse DNS scan, for example).
One could also validate e-mail addys this way, but once again, a simple script can do the same with batches of questionable addys, and one need not send anything to the victims unless one wishes. It really seems as if tracking is all this technique is useful for.