Copyright Infringement Trolls: An “Appreciation of the Situation”

GTA Linux User’s Group, Political Outreach Group (David Collier-Brown, Editor)

Slaw readers may have noticed a flurry of interest around Voltage v. Doe and TekSavvy, an effort to get the courts to compel TekSavvy to identify 2000 households which Voltage Pictures argues are engaging in commercial copyright infringement via the “bit torrent” program.

Slaw has considered the subject of discovery of IP addresses recently in two posts: and

The technical community is following such cases with both interest and trepidation, as we expected suits against individuals to be rare, and limited to $100 to $5,000 statutory damages by the recently enacted Bill C-11. Instead we see new suits, possibly including this one, from what the Samuelson-Glushko Canadian Internet Policy and Public Interest Clinic (CIPPIC) at the University of Ottawa calls copyright “trolls”:

… Trolls’ business model involves alleging that consumers are liable for copyright infringement, and demanding compensation under threat of litigation. The compensation demanded invariably grossly exceeds the damages a troll might expect if the troll were to actually litigate and obtain judgment and a damages award. However, such compensation does not typically exceed the cost to a defendant of defending the action. Enough defendants will choose to pay rather than defend to make the scheme profitable to the troll. The troll typically never litigates through to a judgment, since the costs of doing so would render the scheme as a whole less profitable. The troll’s business model, thus, is an arbitrage game, exploiting judicial resources to leverage defendants’ fear and the costs of defending into a revenue stream.

I and various members of the GTA Linux Users’ Group board are concerned with two different aspects of this. The first is how such a scheme flies in the face of the recent changes to the Copyright Act. The second, closer to home, is the ease with which a troll could use the courts to obtain the contact information of anyone in our community who has used bit torrent as well, but did so to legally download copies of Linux, for example.

The obvious question is what does one do about it if copyright “trolling” breaks out in Canada? The law is not part of our expertise, but, as computer enthusiasts, the facts are. This is therefore an “appreciation of the situation”, to set out the facts for your consideration.

Consider us to be briefing a legislator or an advocate about an apparent new kind of crime. It is up to you, the reader, to decide what the right and wrong of the matter is.


The interested parties are lawmakers, the law societies, lawyers for plaintiffs and lawyers for defendants. Their common aim is to achieve justice. In particular, the aim is to

  • deny commercial infringers the benefits of their crime,
  • to punish private infringers appropriately under the new law, and
  • to deny the opportunity for a plaintiff to make a profitable business out of dropping lawsuits and engaging in private justice.


  • In the United States, the very high statutory penalties for breaches of copyright have lead to judgments like $1.92 Million for 24 songs. In turn, this causes considerable fear in the community and, most recently, the suicide of an accused, Aaron Swartz.
  • The opportunity to settle with fearful members of computer community is a business opportunity, and has led has led firms to “troll” for persons wishing to settle. This sense of “trolling” is something done by a fisherman. The more common use is to refer to the fairy-tale creature that lives under a bridge.
  • To discover people to settle with, the trolling companies collect the internet (IP) addresses of people who use file-transfer programs like bit torrent whom they suspect of copyright breach, and then apply to the courts to have them identified, after which the instigating cases are dropped.
  • This process has been observed many times in the US, with suits used to obtain discovery which are then dropped.
  • The factor that most distinguishes this kind of case from the ordinary is that the plaintiffs show a real unwillingness to bring the defendants to trial. Predominantly, the cases are dropped by the plaintiff after the court has collected contract information from an internet service provider (ISP) and provided it to the plaintiff. Some few are kept open, with individuals dismissed with prejudice, and then everyone is dismissed at a later date. In the latter case, it is hard to distinguish trolls from plaintiffs legitimately concerned about misuse of their copyrighted movies.
  • Because the cases do not proceed to court, the plaintiff only has to show a bona fide claim, much weaker than a prima facie case, to gain the accused’s contact information. This means they can afford to be careless about who they accuse, knowing that if they name innocent persons, some number of them will settle anyway.
  • In the U.S., there had been concern expressed about the propriety of this use of the courts, and the courts have responded in various ways, as described below.

For the purposes of this appreciation, we assume that the trolling process the CPICC describes is undesirable. However, see course 0, immediately below, where we reconsider this assertion.

Courses of Action

Appreciations conventionally start with “do nothing”, and end with something possible but unlikely, to ensure all alternatives are covered. Each course should address at least one factor and the aims of all the parties.
0. Do nothing.

To be fair, one needs to consider the possibility that this what is supposed to happen. A plaintiff may be entitled to use the courts to identify possible infringers and offer to settle with them out of court. An honest plaintiff could well bring a suit to identify apparent infringers, settle with some and bring the remainder to trial. Any harm done by less honest plaintiffs is simply a cost of ensuring the rights of the honest ones. This addresses the needs of all the interested parties, although it will not please some.

1. Break up the cases before discovery.

This is a remedy applied in the United States to ensure that the courts are not used to authorize discovery for cases that will not be heard. Breaking up the cases requires proper per-case fees be paid. It has the effect of disabling the business model of the “trolls”, but does not prevent an action by a plaintiff who sincerely plans to go to trial. This course of action is to the advantage of the administration of justice, the courts themselves, and the defense. It also avoids the need for lawmakers or the law society having to address the problem.

2. Require the case proceed with all the defendants involved.

This would make it resemble a kind of class action suit, with the plaintiff suing a large number of defendants, all with substantially identical facts being considered. This addresses the needs of the courts and the need to do justice to both the plaintiffs and the defendants. It is not obvious, however, if the facts will be similar between accused persons.

Some cases in the U.S. have proceeded with a single defendant at the direction of the court. We do not know the advantage sought here, which suggests we may be missing a factor that is relevant to the discussion.

3. Limit the use of any information disclosed.

The courts or legislature could limit the use of the information discovered to the proceedings in the particular case alone. This would require prohibiting the plaintiffs from communicating with the identified persons except for the case, under the broad supervision of the court. This would have an effect very similar to (2), in that legitimate cases will go forward, but trolls’ use of the contact information for out-of-court settlement will be prevented.

4. Refer the cases to small claims court.

This is to apply once the cases have been severed, discovery is done, and the cases are set to proceed. An honest plaintiff will take all of the above steps, but the so-called trolls only use the courts to impose discovery on the ISP, and have no intention to proceed further. This is only good for honest plaintiffs. If the assumptions are changed such that discovery of the ISP cannot proceed unless the severed cases are committed to be tried, however, it becomes a special case of (1)

5. Refer the cases for Alternative Dispute Resolution.

If the cases, either joined or individually, were referred to alternative dispute resolution, the picture changes. An honest plaintiff can recover at moderate cost, and the defendants will see justice being done, and be able to represent themselves in a less frightening proceeding. A dishonest plaintiff, one who merely desires to frighten possibly innocent defendants into settling out of court, will lose a significant amount of their power to frighten the defendants. This would be attractive to honest plaintiffs, the defendants and the courts. It is not obvious if this would be pleasing to the ADR organizer.

6. Prohibit a particular practice.

Just like barratry, champerty and maintenance of old, legislators could prohibit plaintiff’s lawyers from taking particular actions. In this case, the prohibition would be on using a suit to obtain identifying information which one uses for any other purpose. This would have to be carefully defined, however, to allow honest plaintiffs to proceed. Developing such a prohibition would be an “interesting” task for a legal draftsperson. Deciding what to call it might be even harder: engaging in “troglodytarum”, perhaps?

7. Consider the specific undesirable practice professional misconduct.

The Law Society could conceivably declare this kind of use of the courts to be misconduct before it becomes prevalent, or order a disciplinary proceeding to evaluate it if it is found to have occurred. This would ensure that an honest plaintiff’s lawyer would be able to defend a proper use of the courts, but would be reluctant to proceed if they were acting for a dishonest plaintiff.

8. Rewrite the copyright law.

If the law were to prohibit trolling, dishonest plaintiffs would not be able to use massive copyright suits to force identification of people to threaten. In principle, this could apply to all forms of suits, so patent, trademark and other schemes which misuse the courts as part of their business plans would no longer be possible.

The obvious question would be how? The simplest approach would be a prohibition on joining more than one non-related defendant into a single suit. Of course, that would pose an insurmountable problem to someone who really was being harmed by a large number of people such as “Anonymous”, the internet vigilante group.

It might be better to have a stringent rule about when one is entitled to sue substantial numbers of people, or a requirement such suits cannot be dropped by the plaintiff. This would be an interesting problem to propose to the legal draftspersons among the readership, much like course (6).


… are left to the reader.

Readers with more knowledge of the matters or the law will see additional courses of action and factors affecting them. This is good, as appreciations are a tool to identify broad collections relevant of factors and courses of action that follow from them. Famously, appreciations are used by the military to teach junior officers “how to use their heads for more than hat-racks”.

Readers can contribute to an appreciation by thinking of additional courses and proposing factors that apply to them, or conversely proposing factors and from them creating additional courses of action. The missing factor for proceeding with a single defendant is such an opportunity: the person who contributes the reason can be bringing information that affects all the other courses of action.


  1. David Collier-Brown

    A slightly-distant related article: IBM’s Cheif Patent Counsel on “Patent Trolls”, and suggestions for mitigation…

    His suggestions include
    – Limits on e-discovery
    – Early particularization of issues
    – Early valuation
    – More liberal use of fee awards