Copy protection/DRM is an interesting situation. Let's compare it to encrypted communication.
In encrypted communication, there's users Alice and Bob who want to communicate, and attacker Charlie who wants to listen and/or manipulate. To keep their communication secret, Alice and Bob choose a symmetric cipher to encrypt the data stream. To do this, they need a shared key. They need to find some way to exchange the keys without Charlie being able to listen in. If Charlie gets the key, the encryption is completely useless.
Alice and Bob have various means of exchanging keys. They could meet in person, for example. Or they can use an asymmetric cipher to encrypt the shared key and just publish their public keys. Once the exchange is done, however, they have the key in a secret location and Charlie cannot possibly hope to get it, short of breaking into their houses. The algorithm can be completely open. That doesn't matter, because it is cryptographically secure. As long as the key is safe, the communication is safe.
In DRM, Alice is the content producer, Bob is the media playing chain (from decoder to screen), and Charlie is the pirate who wants to steal the movie. The communication is the media data. This is a highly absurd situation. To decrypt the media, Bob (the media player) needs to have the key. But Bob is a local program on Charlie's local computer. Under normal circumstances, Charlie has full access to everything on his computer. (So-called "trusted" computing means that this is not the case.) This includes the media player, which somewhere, somehow stores the decryption key. DRM's algorithms are open. AACS uses AES, which is well-known. The weakness is the key. Media players go to absurd lengths to hide the key from the user. Obviously, were the players open-source, there'd be no hope of hiding the key. As it is, it's still impossible, but Vista's funky protection features and the length to which developers go to obfuscate the key mean that it's considerably harder to find it.
