The DC Circuit Court of Appeals heard argument today in AF Holdings v. Does 1-1058, one of the few mass copyright cases to reach an appellate court, and the first to specifically raise the fundamental procedural problems that tilt the playing field firmly against the Doe Defendants. The appeal was brought by several internet service providers (Verizon, Comcast, AT&T and affiliates), with amicus support from EFF, the ACLU, the ACLU of the Nation's Capitol, Public Citizen, and Public Knowledge. On the other side: notorious copyright troll Prenda Law.
Copyright trolls like Prenda want to be able to sue thousands of people at once in the same court – even if those defendants have no connection to the venue or each other. The troll asks the court to let it quickly collect hundreds of customer names from ISPs. It then shakes those people down for settlements. These Doe defendants have a strong incentive to pay nuisance settlements rather than travel to a distant forum to defend themselves. The copyright troll business model relies on this unbalanced playing field.
In this case, Prenda sued 1058 Does (anonymous defendants identified only by an IP address) in federal district court in the District of Columbia. It then issued subpoenas demanding that ISPs identify the names of these customers. The ISPs objected to this request arguing that most of the IP addresses were associated with computers located outside of the court's jurisdiction. The ISPs and EFF also showed that Prenda could have used simple geolocation tools to determine the same thing. And we explained that joining together 1000+ subscribers in one lawsuit was fundamentally unfair and improper under the rules governing when defendants can be sued together (known as ‘joinder’).
Unfortunately, the district court did not agree, holding that any consideration of joinder and jurisdiction was "premature." In other words, the court can't consider whether the process is unfair unless and until a Doe comes to the court to raise the issue. By then, of course, it is too late; the subscribers will have already received threatening letters and, in many cases, be reluctant to take on the burden of defending themselves in a far away location.
We believe this ruling was fundamentally wrong. As we've said many times, plaintiffs have every right to go to court to enforce their rights. But they must play by the same litigation rules that everyone else has to follow. To get early discovery, plaintiffs must have a good-faith belief that jurisdiction and joinder are proper. Given the evidence presented to the district court, there is no way Prenda could have formed this good faith belief. So its demand for customer information should have been denied.
The ISPs appealed the district court’s troubling ruling. At the hearing today, the appellate court was particularly interested in the issue of joinder. The court seemed immediately skeptical of the notion of suing 1000 people at once, but wondered if it might be acceptable join together 20 Bittorrent users who had joined the same swarm to acquire the same work. The ISPs and amici said generally no, because the plaintiff can't know whether a given Doe 1 acquired anything from a given Doe 2 – in other words, they aren't necessarily part of the same "transaction or occurrence." We analogized a bittorrent swarm to a casino poker table: over the course of a weekend, a week, or a month, players may come and go, adding and subtracting from the pot, but the players on day one are unlikely to be related to the players on day 4, or day 30.
The ISPs and amici also stressed the issue of burden. While the ISPs were focused on the burden they faced in responding to the subpoenas, EFF directed the court's attention to the fundamental burden on the IP subscribers, noting that the subscribers identified as a result of a subpoena aren't necessarily going to be responsible for any unauthorized activity. An IP address, we explained, only tells you the name on the bill, not who is using the account. In this context, it is crucial that courts attend to the burden on the Does, as well as the ISPs.
The court had a number of question regarding jurisdiction, and directed many of them to counsel for AF Holdings, Paul Duffy. At root, the court seemed to want to know why AF Holdings had not used geolocation tools to help determine where its targets might be located, and why it had not dropped its effort to pursue many of them when the ISPs explained that the Does just weren't in the court's jurisdiction. Finally, the court had some questions about AF Holdings litigation tactics, including the shenanigans that have been widely reported elsewhere.
It is difficult to predict how a court will rule based only on a hearing. But we are encouraged that the judges asked the important and thoughtful questions, and clearly understood both the context and implications of their decision. Many district courts have now concluded that the copyright troll business model is fundamentally unfair, and have taken steps to ensure the judicial process is not abused to foster a shakedown scheme. Let's hope they will soon be joined by the DC Circuit Court of Appeals.