Crypto-Wars: A brief thought on the coming struggle for a neutral net

The weight of network neutrality support is behind a definition of "network neutrality" that might lead to an escalation of end-user cryptographic use. This may or may not be a good thing.

The prevalent definition of "network neutrality" would allow for some forms of discrimination. This thin-edge-of -the-wedge - the proverbial "camel's nose" - would allow for discrimination where IP packets are part of a VoIP application's communication, an IPTV application's communication, etc.; i.e. discrimination for particular service types. In general, this "network neutrality" follows the path described by Lessig and Wu. The Dorgan-Snowe bill appears to takes this approach (12(a)(4) and 12(a)(5) are the relevant sections).

Sidenote: No one appears to take the testimony of Gary Bachula on the red-herringness of QoS seriously . . . Why? (more on this anon)

So, to adopt Wu's language, some discrimination is good, some bad (a sort-of affirmative action policy for the Internet?). What might such a policy beget?

Crystal ball time.

Here's a possible scenario. Recognizing that VoIP packets are prioritized over others, an enterprising software developer writes a bit of code to send ALL traffic as prioritized VoIP packets. Perhaps even sending some useless traffic on the "dirt-road" broadband to avoid attention. To avoid deep-packet inspection he may also encrypt the traffic and/or drop the relevant data (PDF) steganographically into something that genuinely looks like a voice conversation. The network operator will likely respond with more intrusive and sophisticated packet inspection.

A downside. The upshot of this oneupmanship would be an escalation of cryto-warfare that might ultimately prove more damaging to the efficient use of network bandwidth than implementing QoS enables.

An upside. The public might finally (and inadvertently) start adopting encryption tools on a massive scale, leading to greater privacy protections.