It's a Matter of Trust

  • by , February 19, 2007

I’m taking a philosophy class this semester. While it certainly requires a special brand of masochism to willfully subject myself to futile religious debate and a preponderance of ill-referenced Nietzche quotations, what I’m really learning is that one of the core concerns of philosophy is semantics. I repeatedly imagine my professor saying – with much exclamation in his thick, yet endearing, African accent – “What EX-ACT-LY do you mean?” Words are amazing in their lack of inherent symbolic value – without careful examination of semantics, anything we say can and probably does have all sorts of unintended meanings.

When this happens by accident, it’s a double entendre. When it’s intentional, I call it deception. This is how I feel about Trusted Computing. Very quietly, and apparently benignly, the Trusted Computing Group is creating a whole family of systems that have the potential to profoundly affect the digital world, but the names of which make people warm all over – kind of like peeing your pants. Once that initial warmth wears off and you start to think about it, you just get cold and uncomfortable.



I will not claim to be an expert on Trusted Computing. I was first introduced to the concept in a Computer Science class over a year ago. My professor, who of course knows much more about both computers and security than I ever want to learn, spoke with more passion than should ever be allowed in a 9 AM class. Following his introduction, I became very interested in learning more about it. I tried to wade through the Trusted Computing Group’s web page, but it consists almost entirely of the sort of vague techno-babble you would expect to find in a press release; “This group defines architectural framework, interfaces and metadata necessary to bridge infrastructure gaps.” Perhaps I’m just irritated whenever I see the word “metadata." And a missing comma.

Instead, I turned to Wikipedia and the Free Software Foundation. To grossly oversimplify, Trusted Computing is a chip. The Trusted Computing Module (TPM) is a special chip, probably on your computer’s motherboard, that handles all the security-related tasks on your computer. This includes – but is not limited to – input and output, network communication, and of course Digital Rights Management (DRM). When Trusted Computing is active on a machine, most actions are gazillion-bit hardware encrypted using a code unique to your individual TPM. This all sounds dandy to the casual user.

Both of these sites point out that the name “Trusted Computing” is a clever little semantic trick. Your computer is trusted…but not by you. In this usage, a “trusted” system is one that behaves as expected when performing a given task. It is a system that is predictable. Logically speaking, in order to create a more predictable system, one must sacrifice flexibility. So what this really boils down to is control. Trusted Computing is really a system for controlling the capabilities of your computer, ostensibly to prevent it from doing anything it shouldn’t. True, this could mean keeping it from running a segment of malicious code that it shouldn’t. However, who gets to decide what my computer shouldn’t do? Given that this initiative is supported by major players in the computer industry – Microsoft, Intel, AMD, and Sun, to name a few – it is possible that they could decide. That alone is terrifying.

One of the frustrating things about this system is that everyone will be forced to use it. The group’s website states that final control of the system will always be in control of the user. You can simply turn it off. Oh okay, great, then it’s not a problem, right? However, if you have the system at work, or you get an email from a cousin who’s using it, or if you want to check your bank balance online, the remote computer can require your TPM’s unique encryption to verify your identity. Everyone will have the choice to opt-out, but it will be such a ridiculous hassle that no one will. Wow, that’s pretty clever product design, actually. Trusted Computing also has the potential to wipe out Internet anonymity in the name of security, which has all sorts of nasty implications in terms of freedom of expression. The Free Software Foundation puts forth serious concerns regarding monopoly and anti-trust problems, since software use could require TPM authentication.

This is an extremely powerful technology with the potential for terrible abuse and is not some distant blip on the horizon. It is being implemented as we speak. Hey all you Windows Vista early-adopters; you know that cool Windows BitLocker encryption technology? If you’re using it, then congratulations, you’re helping to start the Trusted Computing Revolution!

Perhaps I’m being overly paranoid, and this whole thing is really for my own good. But when someone says “this is for your own good,” how do you generally feel about it? Maybe it doesn’t bother you. In that case, I want to show you something. Okay, now open your mouth and close your eyes. Here it comes…

I began with some philosophy, and I will end with it. If people feel safe, one of two things is true. Either their safety is a complete illusion, or they have sacrificed so much in the name of safety that they don’t really have anything left to lose. Which case do you think we we’re dealing with now?

1 comment about "It's a Matter of Trust".
Check to receive email when comments are posted.
  1. BIG SWINGING, February 23, 2007 at 2:35 p.m.

    dude, double entendres rock!

    Oh, and I vote their safety is a complete illusion.

Next story loading loading..