Special Report A court case that begins this week will define new boundaries in the relationship between US ISPs and creators, regardless of which way it goes.
Music publisher BMG, part of Bertelsmann, is suing cable giant Cox Communications for abetting copyright infringement. The core of the issue is how the ISP handled heavy infringers, subscribers responsible for downloading tens of thousands of files via the BitTorrent protocol.
The case has already caused a minor earthquake: the judge presiding has deemed that Cox lost vital “safe harbour” protections that ISPs enjoy under the DMCA, or Digital Millennium Copyright Act. No detailed explanation has yet been published.
And that in turn this caused insurance house Lloyds to try to rip up its insurance policy with Cox. The insurance house doesn’t want to be on the hook for the damages that could follow an unfavourable verdict against Cox – so it’s pre-emptively suing the ISP to ensure that they won't claim.
The other Safe Harbour
The DMCA’s “Section 512” safe harbour provisions are confusingly, and some say, misleadingly, named. Doubly so, since the “safe harbour” that’s been in the news so much this autumn refers to something completely different: an ad hoc trade arrangement between the EU and the USA over the export of personal data.
The DMCA’s model 1998 internet legislation, widely copied all over the world, was sensibly intended to remove the risk from well-intentioned actors in the nascent internet business. They would have their risks pegged if they satisfied the legal conditions of being a good corporate citizen, by acting in good faith. The removal of liabilities minimised the risk. Section 512 encoded the protections. Without the provisions, the internet may not have grown so rapidly – investors would have been reluctant to back such high-risk ventures.
But two things need to be remembered before we pick up the story. The section of the law that’s relevant here is 17 U.S. Code § 512 a) which affects ISPs. Unlike Europe, the USA carefully distinguishes between ISPs, search engines, caching services, and user generated content hosting services (which “store material at [the] direction of users”). So a bad result for Cox won’t necessarily affect YouTube, the biggest and most controversial UGC hosting service.
Secondly, the limitations on liabilities aren’t really “safe harbours”, in the sense that any ship on the seas can just sail right in and drop anchor there indefinitely. Qualifying for safe harbour requires an eligibility standard to be met, and that includes, quite specifically:
….a policy that provides for the termination in appropriate circumstances of subscribers and account holders of the service provider’s system or network who are repeat infringers
As one critic (requesting anonymity) summed it up to The Register: “The DMCA was never intended to provide someone who is acting illegally a harbour from liability. It was for protecting innocent people from crippling liabilities. It was not a shield for criminality”.
So it’s not a “get out of jail free” card.
Park here, in our safe harbour
In 2011 major ISPs agreed to sign up to a voluntary scheme, the Copyright Alert System, which is a non-legislative tool in the fight against copyright piracy. This was intended to be fairer than the RIAA’s unpopular litigation onslaught, which didn’t distinguish between casual and heavy-duty infringers.
The agreement also protected a user’s privacy, as their identity would not be disclosed to the complainant unless they had a court warrant. And it would kill speculative “trolls”, since subscribers to participating ISPs knew that legitimate warnings could only come from the ISP itself.
Under the CAS system a persistent infringer would be warned. By the fifth or sixth warning, their ISP could throttle the persistent infringer’s bandwidth, or divert them to a landing page. It was up to the ISP to decide exactly what sanctions to provide. CAS only ensnared Bittorrent users and could be easily circumvented by determined infringers using VPNs. But for rights holders it was a small step forward, and both telecomms and copyright groups hailed it as an example of co-operation.
All but one of the major US ISPs signed up to CAS: the refusenik was Cox. Cox has more than 22m residential subscribers in 18 states, and is the fourth biggest ISP in the USA. Cox argued it already had an automated and human graduated response system that worked perfectly well. As with all ISPs, the subscriber signed an Acceptable Use policy, which if violated, could result in “immediate termination or suspension”. Ten or 12 warnings could trigger this, Cox explained.
Rightsholders then turned to a dedicated copyright cop called Rightscorp Inc. Rightscorp’s business is identifying torrent seeders and sending out “parking ticket” notices, typically requesting $20. The company signed up music publisher clients, and went public in 2013. Rightscorp has yet to turn a profit. Its losses are substantial.
A year ago BMG and RightsCorp filed suit against Cox. BMG and RightsCorp alleged that some of the heavy duty ‘torrenters were seriously hardcore. One had triggered 54,489 notifications in 60 days. Each notification was alleged to have been passed on to Cox. Cox ignored them all, the plaintiffs alleged:
By ignoring the repeat infringement notifications and refusing to terminate internet access for repeat infringers, Cox has made an affirmative decision to contribute to known copyright infringement and to continue reaping the substantial financial benefits in the form of subscription fees and fees for higher bandwidth. Cox's conduct renders it ineligible for safe harbor immunity from copyright liability under the DMCA.
We don’t know whether this is what persuaded Judge O’Grady to strip Cox of Safe Harbour liability protections, but something did. The law is clear enough: if an ISP has no for provision for dealing with repeat offenders, then it’s at risk of losing Safe Harbour.
What gets decided?
The case breaks new ground because the legal term “repeat offender” has never really been fully tested. If the case goes ahead, Cox’s argument is likely to be that RightsCorp was bombarding it with repeat allegations, but not repeat evidence, and it failed to meet a decent evidence standard. In its court filings, Cox refers to Rightscorp’s “technological systems” and even “infringement” in scare quotes, arguing that the allegation is merely that.
Rightscorp couldn’t show the source code which Cox requested. It didn’t use version control, which Cox regards as destroying their own evidence (“egregious spoilation”). For its part, Cox seems to have lost (or misplaced) its DHCP logs, files that can tie a subscriber to their IP address at a given time.
So what constitutes repeat evidence? Only if a dedicated 'torrenter has been through the court system more than once, Cox may argue. Which in reality never happens. (It hardly ever happens once, let alone twice).
Did Congress really intend that an ISP's Safe Harbour defence was so broad that even the most flimsy termination policy, one that hardcore infringers barely notice, would suffice? We already know the answer to that one. Judge Liam O'Grady has already decided no, by decreeing that through its behaviour, Cox has forfeited Safe Harbour.
It appears that he thinks, at least in the case of the Cox Torrenter who downloaded tens of thousands of files, that tens of thousands of notifications was good enough to create obligations on Cox's part.
The East Virginia District is set to hear the case on Friday. Judge O’Grady is also the presiding judge in the extradition proceedings of Kim Dotcom – and he sounds thoroughly fed up with sophistry. ®