Why Intel's DRM strategy is flawed

DTCP-IP can and will be hacked

There is no copy protection technology in existance that has not been hacked. This is a cold, hard fact. Copy protection technology in consumer electronics is flawed, and this is why:

An encryption standard consists of the encryption algorithm itself, the content it is protecting, and then the key to unlock the encrypted content. When you buy a DRM'd piece of content, you are given all three pieces of the DRM pie (because if you weren't, your content would be useless); therefore, you are given all the tools you need to open it right back up again. DRM is fundamentally, architecturally, flawed.

DTCP-IP is no different, based as it is on the above architecture. This is something Intel admits: "The industry is not going to be able to stop any rogue attempt to do this," I was told today. "The attempt is to be fair and to respect digital rights and protect them to the best of our ability. To say that this is so rock solid it will never be broken... that's a silly answer to give."

So, Intel admits that this technology will be hacked given enough time and will. This means that piracy will not be stopped by using DRM through a DTCP-IP conduit. If piracy is not the reason for protecting content, what is the reason we are having DRM thrust upon us? That is something I will return to in my conclusion.

If you create a DTCP-IP system, you can be fined $8m

There are a set of standards that are required for a system provider to label their machine or device as DTCP-IP compliant.

Those standards are that 'Level 1 assets and core functions', such as the device key and the encryption algorithm, have to be resistant to attacks from professional tools, and have to be immune to widely available and specialised tools. 'Level 2 assets' such as output functions have to be immune to widely available tools, also.

If it turns out that you are not resistant/immune, you can face an $8m fine. However, there is no definition of what it means to be resistant or immune. We were told by Alec at Cloakware that "The robustness rules are fairly vague... "

Alec suggested that the robustness rules are there to prevent negligence, not to enable the DTCP-IP standards body to sue the company who's device is hacked first. Anyone who does business knows that anytime you get a 'The contract says this, but don't worry, it doesn't really mean it' situation knows that this is a very, very bad sign of incompetency.

So, if you want to develop a DTCP-IP device, you have to adhere to standards that are not defined or face a large fine.

If you buy a flawed system, or you get a virus, your system could be screwed

If it turns out that the device you bought does not fully meet DTCP-IP criteria after all, or if you are subject to a malicious code attack, your device can be switched off and rendered useless.

Part of the DTCP-IP standard is the ability for the device to call home, over the internet, and let the content provider know what you are doing with the content. If it turns out you're doing something that is not allowed, thanks to a flaw in the system you've paid good money for, you can have your system license for DTCP-IP revoked, making you unable to enjoy your content.

This is designed to stop people hacking their players / systems and then attempting to play protected content.

However, it could equally trap an inadvertant user of software which does that for them: how many spyware installations are there on computers around the world? How hard is it, these days, for someone else to do something to your machine? As an example, a malicious virus could enter your system and break DTCP-IP using whatever inevitable crack becomes available, making your system absolutely useless for playing premium content in the future.

Intel's answer to this is that it is up to your individual system provider to make sure it is not cracked. However, based onthe architecture above, we know that this is impossible. This continues Intel's trend of "Don't shoot us..." If the user is attacked by a malicious code infection, there is no mechanism in DTCP-IP to un-screw the system once it is broken. Intel is promoting the use of a system that can't be fixed if it is broken, and saying that that's not its problem, it's up to the system creator, not them. Don't forget: Intel is just the messenger, right?

Guess what? It really is Intel's problem: a hacked, broken Viiv system does not make for a good end user experience.

DTCP-IP does not support the most popular system on the market

The protocol itself is DRM-agnostic, as we mentioned above. It will take content from anyone and transmit it securely, providing the content provider works to protocol standards.

The biggest provider of secure, legal content right now is Apple, with its iTunes music store. Apple does not make its content compatible with DTCP-IP, meaning that anyone using a Viiv PC cannot securely stream Apple's content to another Viiv-compliant device.

Sure, this is up to Apple to fix, in some respects: but there is no indication forthcoming from Intel that it is even engaging in talks to make this happen and to give users the ability to use their iPods and Apple content with this technology.

Not supporting the most popular music system on the market does not make for a good end user experience.