A security research company has reported that Apple's iTunes software creates vulnerabilities in Windows XP and Mac OS X that allow remote parties to install and execute code on your computer — basically, the vulnerability means that your computer can be taken over by an attacker who can then hijack your connection, your files, and your privacy.
I'm not surprised. An objective of good security is to protect users from attackers who want to prevent the user from controlling her computer. DRM — like that in iTunes — is a system for allowing remote parties (e.g. entertainment companies) to enforce their policy on your computer. Once you design the system to let anyone apart from the owner to control it, you open up the possibility that someone other than the owner will end up controlling it.
As with Sony's rootkit, every DRM has the potential to create this kind of vulnerability. Imagine if Yale manufactured every door-lock so that a "master key" from Yale could open it. So long as no one except Yale knows about the master key, you're safe (assuming you trust Yale). But someone always finds out — that kind of secret is too valuable to remain a secret. Once a bad guy knows that there's a single technique that can be used to access every door with a Yale lock, it's only a matter of time before the attacker develops a crack.
DRM systems are an attractive nuisance, the cracker's best pal. They are, at root, systems for giving control over your computer to someone other than you. That's an invitation to disaster.
It may be that this iTunes vulnerability takes advantage of the auto-update or the connection to music store and not the components where iTunes keeps you from directing your own computer to do your bidding, but having such systems is just a bad idea if you're trying to design systems that protect their owners.
(via /.)
