Why AT&T's Plans To Filter The Internet Will Only Do More Harm To AT&T (And Everyone Else)

from the not-a-good-idea dept

AT&T announced last summer that it was going to start filtering traffic for copyrighted content -- so we're still not entirely sure why many in the press seem to think it was something new when discussed at CES a few weeks ago. However, this new burst of attention has many more people pointing out all the reasons this is bad for AT&T itself. As we said, this seems to make no sense at all, unless it's some bizarre attempt to come up with an excuse to get rid of net neutrality. In that post, we noted that any filtering would likely open up additional liabilities for AT&T, potentially losing its safe harbors from being a service provider (safe harbors that AT&T itself spent a lot of effort lobbying to have put into the law). Tim Wu has a lot more detail on that aspect of this plan (which he calls "corporate seppuku"). However, there are many other problems for AT&T as well. For example, it won't take long for someone to accuse AT&T of violating wiretap laws, a charge which may be accurate. But the biggest point is that this won't even do what they hope it will do. It won't stop unauthorized transfers from happening and it won't reduce network traffic. As we've discussed in the past, every move to do this kind of filtering will only drive up the market for encryption technologies, and that encryption actually adds more overhead to internet traffic. The PC World article linked above notes that 20% of all bittorrent traffic is encrypted, and if that number goes up, as it will under a filtering regime, the network load will only increase. So, if AT&T actually thinks (as it sometimes claims) that filtering will decrease the burden on the network, it's likely very mistaken.

Filed Under: encryption, filtering, liability, network neutrality
Companies: at&t

Reader Comments

Subscribe: RSS

View by: Time | Thread

  1. identicon
    Michael Evans, 22 Jan 2008 @ 12:23pm

    Might be repeating some stuff, but I'm short on time right now and want to make sure these points are addressed.

    1) Encryption is a benefit to anyone wanting to keep private things private, or control distribution of their information. It will be central to business models, private communication, and anyone trying to break the law to any degree... and you won't be able to tell who is doing what. You can't force people to hand over their freedom by making them use a vulnerable algo, and processing for that on the fly would cause insane lag anyway.

    2) When you make the types of content transfered transfer at different rates, without being fair or transparent about it, the easiest solution will be to encapsulate/tunnel the content. Any redundancy or headers already there will then be repeated or even increased in the outer wrapper.

    3) Get a business model that works.

    My ideal system would look something like this for billing.

    * Account Fee: $5 per year
    * Network Action Hotline: Free (Automated front end, you dial in the account number, it tells you what they think the status of the network is for you (up, down, area outage, etc) and if you disagree you can dial in details and checkoff a simple troubleshooting list over the phone. If at that list you don't mind being called back by a technician for additional questions you can leave them a callback number.)
    * Network Support Hotline: $5 flat rate to do the above with a human. (If they can't help you isolate the problem in 30 min, they deserve to eat it.)
    * Data Transfer: Billed by mega-bit (or byte, multiply cost by $ per second rate guarantee... )
    * Data Pipe Class: (Depends on the hardware, flat rate lease per month...)
    * Service Level: Various types of response packages, redundancy options (basically discount extra-links), and insurance against downtime.

    ...which means you're leasing that much of a pipe, dedicated 24/7. Rate structure something like BulkRate*(1+1/(Weight^((Speed-X)/(Y-X)))) , where Speed is the constant BPS, X is the speed where rate is 2x, Y is the speed where rate is close to 1, and Weight is how bent the slope is (If the Weight is too high relative to the difference in 2x and ~1x rates, then odd distortions in pricing occur.)


Add Your Comment

Have a Techdirt Account? Sign in now. Want one? Register here

Subscribe to the Techdirt Daily newsletter

Comment Options:

  • Use markdown. Use plain text.
  • Remember name/email/url (set a cookie)

Follow Techdirt
Insider Shop - Show Your Support!

Report this ad  |  Hide Techdirt ads
Essential Reading
Techdirt Deals
Report this ad  |  Hide Techdirt ads
Techdirt Insider Chat
Report this ad  |  Hide Techdirt ads
Recent Stories
Report this ad  |  Hide Techdirt ads


Email This

This feature is only available to registered users. Register or sign in to use it.