The Bloc targets ISPs, but what are ISPs?

A constituent received a letter from Bloc leader Gilles Duceppe on Bill C-61 that looks quite similar to what we saw from Bloc MP Thierry St-Cyr. The message: we somehow have to fulfill some obligation to ratify the WIPO treaties that the Liberal federal government signed, and the Bloc is worried that ISPs aren't being held more responsible for infringement.

Now here is my personal problem. Bill C-61 is an omnibus bill, with the majority of the bill being about "technological measures". A colleague did a work count of the English text and found that this component of the bill takes up over a third. If we were able to talk about ISP liability as an independent topic, I would have a more nuanced position that lies somewhere between no liability and the Bloc position of scapegoating communications providers.

Bill C-61 uses the word "Internet" about 43 times, but the term is not defined. Those of us following the related Network Neutrality debate know that definitions of terms matter. Is the "Internet" being used to refer to any digital communications network which Canadians can hire to interact with? Is it any digital network designed under the end-to-end principle (Where the endpoints are smart, and the network is dumb by design!)? It is any network that uses the TCP/IP protocols no matter how it is configured?

These differences matter when discussing things such as ISP liability. I believe that a network configured under the end-to-end principle where network operations are simply conduits that do not inspect or modify the contents of any traffic should not be liable in any way for that traffic. If, however, the providers wish to inspect and/or modify the traffic, then they should have liability relative to the intrusive nature of their inspection/modification. Is phone and cable companies want to offer a service they market as the "Internet" that acts more like a phone or cable service, then they should be expected to be liable as phone and broadcasting companies.

The TWIT show Security Now has been doing a series of episodes on ISP privacy and the breakdown of ISP trustworthiness (Example: Security Now 151: Frakking Phorm)). There are companies that want to do demographics on Internet customers by installing equipment in an ISP and then inspecting/modifying traffic such that this information can then be used for direct marketing purposes. The violations of basic network protocols and the privacy of the Internet customer are enormous.

We need to be aware of the thinking of these ISPs and marketing companies when discussing things like ISP liability. A technical person will look at the end-to-end principle that was intended to be the design of the Internet, and think it is wrong to force these folks to inspect and/or modify (filter) traffic. But we need to realize that not all companies marketing services called "Internet access" are actually offering a service we technical people would recognize as an Internet service.

I strongly support the notice-and-notice regime in Bill C-60 and Bill C-61 for companies that are operating as end-to-end network operators of dumb networks, routing packets and never inspecting, filtering, or modifying the contents, or prioritizing based on the contents.

But companies who want to offer something different than a proper Internet connection should not be able to make use of a regime designed for real Internet Service Providers.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

The liability shield is extended to all digital networks

The definition of the term "Internet" is largely irrelevant to the liability shield for service providers under bill C-61. Section 21 grants that exemption to those who provide "services related to the operation of the Internet or another digital network". The second half of that or construction makes the first half meaningless, unless the Internet is somehow not a digital network. All digital networks qualify.

Of the 42 instances of the word "Internet" (not sure why my count doesn't agree with yours, but it probably doesn't matter very much), 8 are in the preamble and unqualified, while 14 have essentially the same qualification: or any other digital network. The remaining instances deal with end-user liabilities or protections, including the educational exemption (12 instances); the provision that contracts can negate format-shifting rights on downloaded items (4 instances); and, my personal favorite, the oddball restriction on time-shifting material downloaded from the Internet (4 instances).

All digital networks qualify.

Thanks for the update. Also not sure why the differences in word count, but then again I did put this one together as a rant about some of the oddball things some "ISPs" are doing. I was too lazy to figure out how to do a proper word count, so I just did a search-and-replace and got the number from that ;-)

I'm curious if you think all digital networks should qualify for the notice-and-notice regime, or whether we should have better ways to differentiate networks that should qualify from those who should not. Are we basing our notice-and-notice on believing that companies operating as "common carrier" neutral networks should only be expected to carry out the same neutral communications roll in communicating between alleged copyright holder and alleged copyright infringer? What about networks that are operating very differently?

Am I just biased in that I want incentives for businesses that want to offer a neutral dumb network so we don't end up transforming the "Internet" into a thousands of channels cable network with nothing on? I'm quite upset at the types of things I keep hearing about with the inspection and modification that some legacy phone and cable companies want to do.


Free/Libre and Open Source Software (FLOSS) consultant.

notice-and-notice

Well, I would at least be hesitant to link the notice-and-notice regime to the net neutrality question. Some, perhaps most, of the benefits of notice-and-notice accrue to the users: less fear of arbitrary removal or destruction of legitimate content; less risk of invasion of privacy; generally fewer opportunities for harassment by self-proclaimed rights holders. It would seem unfortunate to deny users these benefits because network operators want to mistreat them in other ways.

I can see the appeal of penalizing invasive actions by network operators, or conversely of rewarding neutrality. But this doesn't seem like the right lever to apply. The force is applied in the wrong location.