Steven Price

My book

Media Minefield


Guide to NZ Media Law

Official Information Act

Official Information Act


Bill of Rights Act

Media law resources

Feeds (RSS)


« Officials drop the BORA on 92A | Main | A copyright anomaly »

A post on 92A that David Farrar probably won’t link to

By Steven | February 24, 2009

You may have noticed that I didn’t join the blackout in protest against section 92A.

It’s not that I don’t have concerns about the law. As you can see from the post below, it doesn’t seem to me that it’s been well thought through. I’m open to the idea that it goes too far.

But “guilt by accusation”? There’s nothing in the law that requires that. Here’s section 92A:

92A Internet service provider must have policy for terminating accounts of repeat infringers

(1) An Internet service provider must adopt and reasonably implement a policy that provides for termination, in appropriate circumstances, of the account with that Internet service provider of a repeat infringer.

(2) In subsection (1), repeat infringer means a person who repeatedly infringes the copyright in a work by using 1 or more of the Internet services of the Internet service provider to do a restricted act without the consent of the copyright owner.

An ISP that simply terminated on the basis of two or three allegations would surely not be implementing its policy reasonably, and termination would be inappropriate. True, this provision is awfully vague (which may also raise Bill of Rights issues, actually). Yes, it gives quite a lot of wiggle room to ISPs (though they are not likely to try to implement it against their clients’ interests). But it does not require - or even permit - guilt by mere accusation. [Dean Knight has made the same point].

Some people are getting het up about the fact that an account may be terminated without the protections (and standards of evidence) available in a court. But is that so odd? Your phone company can cut you off for breaching its policies. Your employer can fire you for breaching policies. Your ISP probably has other policies reflected in its contract with you allowing it to terminate your account if you peddle kiddie porn, for example. If push comes to shove, you can take the issue to court. As you can do under section 92A. But the development of policies (even mandatory ones) that have real consequences for breach without any necessary court involvement is nothing new.

Throw in the new Telecommunication Carriers’ Forum draft on how this law is likely to work in practice and we move even further away from guilt by accusation. What this draft gives us is “innocence by denial”. Under this plan, a copyright holder who asserts breach triggers an “education notice” to the user, who can simply deny the breach by issuing a counter-notice and no further action needs to be taken by the ISP. Really. There is a provision in there saying that ISPs can, in their discretion, decide to take action if they think the denial process is being abused. But how many ISPs are going to do that?

The draft  is worse than that, from the copyright holders’ point of view. A copyright holder has to pay a fee before the ISP even has to look into the allegation. So if a copyright holder can provide crystal clear evidence that a user is downloading files illegally, the ISP won’t even glance at it unless the holder can stump up with some dosh. I have some sympathy with the ISPs’ desire for cost recovery, but I’m not sure that any policy that allowed an ISP to overlook incontrovertible proof just because someone hasn’t paid it some money could be regarded as reasonable. (I’m not saying proof will always be incontravertible, I’m just mentioning the easy case for the purposes of argument).

Actually, that raises another question: where the hell do the copyright holders get their evidence that some particular person is infringing? Beats me. But I do note that the draft puts that evidential threshold pretty high.

Does this draft really comply with section 92A?

The copyright holders are proposing a slightly amended system - in which they get to look at the users’ reasons for denying infringement via their own resolution process, and if they end up still thinking there’s been a breach, the ISP has to accept that. But that doesn’t seem fair either [and, as someone has pointed out in the comments section, that comes closer to guilt by accusation].

I think any fair and reasonable system must involve a quick and fairly cheap disputes resolution body. And I note that we already have a Copyright Tribunal set up and functioning, though for some bizarre reason, all it does is rule on issues concerning licensing systems. I can’t help thinking that it would be a good idea to broaden its jurisdiction to include ISP copyright issues, set up a lower-tier fast-track mechanism for resolving the 98% of infringement claims that are slam dunks, and refer the tougher ones (eg where there’s a fair dealing issue) to the Copyright Tribunal.

[Update: credit to DPF, he did link to this post. He says he disagrees that section 92A isn’t guilt by association, but doesn’t explain why. On the other hand, he has written a lot on this issue, and you can see his thinking there.]

Topics: Copyright, Internet issues, NZ Bill of Rights Act |

7 Responses to “A post on 92A that David Farrar probably won’t link to”

  1. Andrew Easterbrook Says:
    February 24th, 2009 at 2:27 pm

    The only point I have to add is re this bit:

    > Actually, that raises another question: where the hell do the copyright holders get their evidence that some particular person is infringing? Beats me. But I do note that the draft puts that evidential threshold pretty high.

    Evidence gathering is pretty easy - a recording company or similar would upload a fake torrent or participate in the p2p networks. This gives them access to a random selection of IP addresses (those users that happen to get part of the file from the company’s “planted” host computer). Filter the IP addresses to only NZ ones and you have date + time + file size/name/description + IP address, which is enough to go to an ISP and file a notice. The code prohibits the ISPs from releasing names, as far as I can tell, so they act as an intermediary, thus preventing any possible breach of privacy issue.

    The problem is that date + time + file size + IP address is only proof *something* has occurred - the ISP can’t, in the more tricky cases (eg me composing a song very similar to a 50 cent song and uploading/downloading it) really be said to be in a position to decide whether copyright has been infringed or not. So it’s left to the copyright holder to assert that copyright has been infringed in that particular case, via a copyright holder notice, and the code of conduct takes this as an example of evidence that would satisfy a court (cl 11.2).

  2. lyndon Says:
    February 24th, 2009 at 2:43 pm

    I think my explanation for the position people have taken is that the law being interpreted reasonably isn’t what they’re afraid of. There’s kind of a mash of systemic issues (like ISPs being more afraid of RIANZ than any given customer) and what the recording industry will do, or try on, with a law overseas. I feel the conculsion is not unsupported.

    And I’d note RIANZ’s position appeared to be that it was or should be a guilt-on-accusation regime.

    (Of course, I don’t like anything else about the clause either. I’m not anti-copyright but the internet works by making copies of things and you need to choose your battles. I think what you might call the industry approach is impractical, counterproductive and keen to be disproportionate.)

    Incidentally it’s been noted that for most ISPs illegal downloading probably does breach policy. Which no doubt has implication for both sides.

  3. mushion22 Says:
    February 24th, 2009 at 2:51 pm

    The process used for detecting copyright infringement and it’s downfalls is detailed here http://dmca.cs.washington.edu/

    Indeed, the law itself under S92A does not stipulate guilt on accusation, but it has given uncertainty and risk, and thus leads to policies (read: TCF CoP) that include internet accounts being disconnected upon accusations.

    While contractually this is legitimate assuming the terms of sale/service include this policy, the debate should be centering around whether this is an appropriate measure to combat piracy and whether ISPs should be in a position where they need to disconnect customers without a court order in order to protect themselves from liability. (Based on the premise that ISPs prefer to keep customers even if they infringe copyright - ie ISPs prefer to remain a transport mechanism just like the post office but cannot do so because of exposure to liability for copyright infringement).

  4. Michael Says:
    February 24th, 2009 at 3:05 pm

    I don’t think that’s accurate. The ISP will *never* “look into” anything, because a) they can’t (it’s just not technically possible in any reasonable fashion*), and b) they aren’t positioned, even if they had that information, to make any sort of judgment about it.

    The COP as written is really just a buck-pass - send as much as possible downstream, by defining “ISP” as broadly as possible, and try not to get sued on the rest.

    The “pre-approved copyright holder” distinctions it draws are downright weird too. Why do non-approved holders get less information back about downstream ISPs? They’re the only ones who have to pay in advance too, by the way.

    The supplemental counter-notice strawman proposal in the draft from RIANZ/AMCOS/etc absolutely is guilt by accusation. Even without it, the “don’t get sued” impetus for ISPs makes it amount to the same thing, unless and until they can convince a court that their COP is adequate. You can hardly expect them to go to the wall on behalf of one of their users.

    * This is a point that hasn’t been explored enough. It’s utterly impossible to evaluate that sort of accusation unless you record *all* the traffic that you handle, for every customer, and retain that for at the least a number of weeks. That adds up fast, and it’s a substantial burden on the ISP (if it’s even legal). Parsing the log afterwards is hard. If the traffic is encrypted, it’s nearly impossible. If the “copyright holders” really want to evaluate claims on their merits, it’s going to be very expensive and not very effective. I really don’t see a way around that.

  5. Blog Bits | Kiwiblog Says:
    March 1st, 2009 at 7:37 pm

    […] Steven Price says that s92A did not mean guilt by accusation. I disagree with him, but agree that the solution should be some sort of dispute resolution scheme. […]

  6. Graeme Edgeler Says:
    March 2nd, 2009 at 12:06 am

    Update: credit to DPF, he did link to this post…

    Surely it was a challenge he couldn’t refuse :-)

  7. Rick Shera Says:
    March 6th, 2009 at 2:26 pm

    Interesting comments here, which are hopefully being put into submissions to the TCF.

    Re the “guilt by accusation” label, that I think did a job of getting public attention even if s92A itself says nothing of the sort. However, I think the label is shorthand for the practical consequences of rightsholders’ public position on repeat infringement - the so called three strikes, graduated response process. When iiNet in Australia decided not to adopt such a policy but instead referred all complaints to the police, it was (and still is being) sued by 34 TV and film companies who effectively are saying that an ISP should accept their evidence at face value. Yes, there will be plenty of open and shut cases but there will also be cases of technical mistakes in IP addressing, false claims by competitors, over zealous restriction on fair dealing rights, copyright law being used to stifle free speech, contract disputes dressed up as copyright claims and even false claims to ownership (e.g., where the material has fallen into the public domain). ISPs are just not equipped to decide on these and yet need a policy that caters for all allegations.

    Expanding the Copyright Tribunal’s jurisdiction or using something like the .nz domain name dispute resolution process is the answer I think - but who pays?

Comments

You must be logged in to post a comment.