How much control can — or should — communications transport providers exert over the content they carry?
Some background: Back in the pre-Revolutionary U.S., transport and content providers were one and the same. The folks who wrote newspapers also ran the mail services — which meant that they’d never bother to deliver the competition’s newspapers.
One of the benefits of establishing and regulating the U.S. Postal Service was creating a legal framework prohibiting carriers from interfering in content.
Over time, the principle of separation of content and transport became established, not just for mail but also for telephony — the phone company can’t censor your calls, or even wiretap them, without an explicit request from law enforcement. And so on.
You’d think that the principle would continue to apply in the Internet age. You’d be wrong.
For a range of reasons, ISPs have increasingly opted to tinker with the content they transport. Here are a few examples:
• Back in March, VoIP provider Vonage complained to the U.S. Federal Communications Commission about the fact that broadband providers, including cable and wireless companies, were intentionally blocking Vonage’s VoIP calls. The FCC subsequently fined one of the broadband providers and earlier this month issued a policy statement condemning the practice. Left undecided was the question of what happens when a provider doesn’t block traffic but merely drops it to a lower level of QoS (see below).
• Earlier this summer, Canadian telco Telus blocked its 1 million Internet subscribers’ access to a pro-union Web site (Telus was then in dispute with its union.) The Canadian government, including the Canadian Radio-television and Telecommunications Commission (CRTC), is reviewing the case and may revise its regulations to more strictly prohibit the practice.
• Many service providers, including Sprint, MCI, AT&T and AOL, offer services such as malware-blocking and distributed denial-of-service (DDoS) prevention in which the carrier will proactively drop content that its subscribers consider objectionable, such as spam, viruses, spyware and, potentially, peer-to-peer traffic.
Should the separation of content and transport be maintained? As a basic principle, I’d say yes. Hats off to the FCC and (hopefully soon) to the CRTC for coming down hard on providers that flagrantly violate subscribers’ best interests.
But sometimes having service providers tinker with content is a good thing. Many subscribers (who, it should be noted, explicitly request — and usually pay extra for — the services) consider network-based malware-blocking and DDoS prevention benefits.
Things get less black-and-white when you introduce issues such as QoS.
Should a carrier be forced to transport a competitor’s voice services at the highest available priority? Or is it OK to drop the packets to “best-effort” priority unless the competitor pays for top-priority QoS?
I’m inclined to say that in this case bumping the packet priority is fine — so long as the VoIP provider has the option of purchasing (and paying for) top-tier QoS.
Bottom line: The separation of content and transport is a great general principle — but sometimes it’s worth a closer look.
QuickLink: 058175
–Johnson is president and chief research officer at Nemertes Research, a leading independent technology research firm. She can be reached at johna@nemertes.com.