The FCC’s case against net neutrality rests on a deliberate misrepresentation of how the internet works

The FCC has just published the notice of proposed rulemaking that would roll back the 2015 Open Internet Order establishing net neutrality. Its first and primary justification for doing this is a way of defining broadband access that’s so backwards it’s ridiculous. It would be funny, if the future of the internet didn’t depend on this incredibly disingenuous maneuvering.

After the introductory bloviation and half-told history, the “Restoring Internet Freedom” proposal begins its argument in earnest. The first point the FCC makes is regarding the text of the 1996 Telecommunications Act, and how it defines “telecommunications service” (how broadband is currently defined, allowing net neutrality rules to be effected) and “information service” (how it was before the net neutrality rule).

Now, I’m going to list the two definitions. Which one do you think sounds like what a broadband provider does?

  1. “The offering of a capability for generating, acquiring, storing, transforming, processing, retrieving, utilizing, or making available information via telecommunications, and includes electronic publishing, but does not include any use of any such capability for the management, control, or operation of a telecommunications system.”
  2. “The transmission, between or among points specified by the user, of information of the user’s choosing, without change in the form or content of the information as sent and received.”

Take your time.

Okay. Number 2, right? Because your ISP doesn’t store the data you post on Facebook, or the address you look up on Google Maps, or the Pope you read about on Wikipedia. It’s edge providers like the ones I just mentioned that do all the “generating, acquiring, storing,” and so on. ISPs just transmit the information, don’t they?

Perhaps it would surprise you, then, to hear that the FCC has the exact opposite idea of how the internet works! The following (don’t skip it) is from paragraph 26 of the new order.

Whether posting on social media or drafting a blog, a broadband Internet user is able to generate and make available information online. Whether reading a newspaper’s website or browsing the results from a search engine, a broadband Internet user is able to acquire and retrieve information online… In short, broadband Internet access service appears to offer its users the “capability” to perform each and every one of the functions listed in the definition — and accordingly appears to be an information service by the definition. We seek comment on this analysis.

Oh, you’re going to get comment on that analysis, all right. Let’s just run down the obvious objections.

First, most broadband providers simply don’t offer the services listed. Does Comcast have a blogging platform, a news service or a search engine that anyone actually uses on purpose? No, Comcast offers a pipe through which to reach other services. Reaching other services is the service that it itself offers! There are even safe harbor laws shielding it from liability for transmitting the things provided by and through those services — because ISPs have no control over them!

Second, broadband providers often aren’t even aware what information they are transmitting, because it is encrypted. They cannot transform it, generate it or process it because tampering with encrypted data is immediately evident to the users. While “drafting a blog,” as I’m doing right now, the ISP only knows that it is handing packets between the router at this coffee shop and the IP address/port belonging to WordPress. It doesn’t know I’m publishing something contrary to its interests. Sorry, ISP.

Third, most services that are in fact offered by the ISP, such as DNS lookup, error pages, caching and routing, all have to do with reasonable network management — the work of getting packets from one place to another properly. That class of things has a big cutout in the net neutrality rules because the internet is a complex place with lots of moving parts.

This analysis is like saying that because someone built a bridge, they also created the entire city and all its resources on far side of it. It’s absurd, and in fact the argument was already tried to little effect in a federal court just three weeks ago. Anyone with a modicum of technical knowledge will find this explanation of how the internet and web work truly wrongheaded and entirely incorrect. It’s hard to think of this as anything other than a willful misrepresentation of the facts.

But wait, there’s more! Not only does the FCC’s proposal claim that broadband providers do things they don’t, it claims that they don’t do things that they do!

Let me explain. Specifically, the order denies that ISPs provide telecommunications, as defined above in No. 2. (That’s not an antagonistic paraphrase: the quote from paragraph 29 is “Internet service providers do not appear to offer ‘telecommunications.'”)

For one, broadband internet users do not typically specify the “points” between and among which information is sent online. Instead, routing decisions are based on the architecture of the network, not on consumers’ instructions, and consumers are often unaware of where online content is stored… Even IP addresses may not specify where information is transmitted to or from because caching servers store and serve popular information to reduce network loads…We believe that consumers want and pay for these functionalities that go beyond mere transmission — and that they have come to expect them as part and parcel of broadband Internet access. We seek comment on our analysis.

Well then, since you asked!

First, users absolutely, 100 percent do specify the points between and among which information is sent. They specify it most commonly by using a “uniform resource locator” or URL, the shorthand used on the entire worldwide web to describe where a given item is stored. Users request these specifically, and where they aren’t obvious (such as with a generated URL, or content hidden behind a web app), the “point” the user is requesting is still implicit in the interface and protocol used. Seriously: The entire internet and web are built on the fundamental idea that users specify these points, and the transfer protocols figure out the rest.

Second, if you want to be realistic about the definitions here, “point” could just as easily indicate a service or company, like Netflix, as it does an associated IP address or server rack. We don’t know the exact desk at the IRS to which we mail our taxes; the “points” we choose are our mailbox or carrier and the all-purpose address or inbox of the second party. The IRS is synonymous with the address it lists, and Facebook is a point just as much as is. The FCC specifically addressed this issue in the 2015 Order (on paragraph 361).

Third, “mere transmission” is an exact description of the role played by ISPs in the web and internet ecosystem. There’s really nothing “mere” about it, though — “merely” transmitting high-definition video at several megabytes per second across hundreds of miles or more is a titanic feat requiring much infrastructure and expertise. But it’s what ISPs are good at, and it’s what we pay them for.

Fourth, users of landline telephones never knew the exact locations of every switch, operator, relay and transfer station. They did indeed rely on telephone companies to supply this information — as part and parcel of the telecommunications service they asked for.

Fifth, for that matter, the ISPs themselves don’t know the location of most of this data. As I mentioned before, much of it is encrypted, or passes through a proxy, or is cached on the fly by Google or Netflix. ISPs follow the instructions of these services as items are shifted among the global data centers run by Pinterest and Instagram, dutifully transmitting packets from the point specified by the user to another point, either themselves or another service.

These are the problems I encountered in the first two pages of this section. The intro is also full of misinformation, most of which, however, I already addressed in this article summarizing the arguments against net neutrality.

The NPRM is 75 pages, and I plan to read every word of it (Update: I have, and you should too! It’s perfectly intelligible). But I had to get this astonishingly wrong beginning out into the open, because boy is this a bad start for the FCC’s agenda.

Want to let the FCC know about how wrong their analysis is? Here’s how to comment on the proposal to roll back net neutrality.