On Monday, the Supreme Court is going to hear oral argument in a significant copyright case, Cox Communications v. Sony Music Entertainment. The issue before the Court is the extent of contributory copyright infringement liability, something the Court has considered twice in recent decades, in the famous Betamax case (Sony v. Universal) in 1984, and in MGM v. Grokster in 2005.

I’m interested in almost any appellate case on copyright law, but I was interested enough in this one that I submitted an amicus brief to the Court arguing how it should come out. This post will introduce the dispute in Cox case and how it emerges from the history of contributory liability; tomorrow I’ll explain how the Supreme Court’s prior intervention in Grokster has added to the doctrinal confusion; and finally on Sunday I’ll explain why I decided to take the time to write an amicus brief. Hopefully on Monday I’ll have time to do a quick review of the argument.

The Cox case represents yet another battle between content owners and technology companies over the extent of indirect liability for copyright infringement, that is, liability internet service providers might have for the infringing acts of their users. For the past two decades, much of that fight has been over the conditional immunity for ISPs provided in 1998’s Digital Millennium Copyright Act, but the Cox case returns the debate to the underlying obligations imposed by copyright law itself: when does an intermediary like Cox have to stop infringers from using its service, and when can it safely regard those infringements as Somebody Else’s Problem?

The legal question here quickly enters some deep policy waters. Intermediary liability is recognized in many areas of the law, from torts to securities fraud to criminal law to all areas of intellectual property. To be effective, intermediary liability needs to strike a careful balance. First, the direct wrongdoers have to be, in some way, difficult to pursue—if they aren’t, then there’s no need to impose liability on someone else. And second, the intermediary has to have both the knowledge and the ability to narrowly target the bad acts without causing unnecessary spillover harms to beneficial activities.

Part of the problem in achieving that balance in the modern era is that the very notion of case-by-base balancing—by courts, by regulators, by almost anyone—has gotten a bad name. As I argued in my recent article The Grapes of Roth, that style of decision-making has faded, replaced by attempts to limit judicial discretion by rigidly following the text of either statutory provisions or multi-part tests.

Recently, however, I thought I detected some inclination by some of the justices to cut back against that trend and instead emphasize that the overall balance of intermediary liability emerges from the interplay of various considerations. So I decided to give that inclination whatever additional nudge I could with my brief.

The case itself is no trifling matter. Cox, a cable company that provides broadband internet access to its subscribers, is appealing a $1 billion jury verdict holding it liable for assisting some of those subscribers in engaging in copyright infringement. The case arose from an effort by record labels and music publishers to stem the tide of peer-to-peer filesharing of music files by sending notices of infringement to access providers such as Cox. The DMCA bars liability for access providers as long as they reasonably implemented a repeat infringer policy. But who’s a repeat infringer? The notices were an effort to get access providers to take action by giving them the knowledge of repeat infringements necessary to trigger the policies.

According to the evidence presented at trial, however, Cox was extremely resistant to receiving or taking action in response to the notices. The most colorful bit of evidence (and yet another example of how loose emails sink ships) was from the head of the Cox abuse and safety team in charge of enforcing user policies, who screamed in a team-wide email: “F the dmca!!!” (This was followed by an email from a higher-level executive on the chain: “Sorry to be Paranoid Panda here, but please stop sending out e-mails saying F the law….”) The Fourth Circuit ultimately held that because Cox didn’t reasonably implement its repeat infringer policy, it lost its statutory immunity, and then at trial the jury found Cox contributorily liable for the infringements that it had been notified about, which the Fourth Circuit affirmed.

The question before the Court is whether the lower courts applied the right test for contributory copyright liability, or applied it correctly. (There’s a second question, about the standard for willfulness in determining damages, but I didn’t address that one.) There’s a couple of things that make this issue difficult to disentangle; one has to do with the history of contributory infringement doctrine, and the other is a technical issue about what, exactly, is being challenged on appeal.

First, the backstory. Contributory infringement liability emerged in patent, copyright, and trademark cases in the late nineteenth century. But courts were not given to identifying a full list of all the necessary and sufficient conditions for such liability. Rather, in true common-law fashion, they would simply recite some facts and then reach a conclusion, such as that “[u]nder such circumstances the defendant … occupies the position of a party acting in concert with the purchaser … , and is responsible with him as a joint tort-feasor.” Harper v Shoppell, 28 F. 613, 615 (C.C.S.D.N.Y. 1886).

That style of analysis works well when efficiency and finality are the important goals of a decision-making system. For example, evidentiary rulings during a trial are often made with a minimum of explanation and limited opportunity for appeal. But by the 1960s observers expected more from published opinions.

In the 1960s, copyright, like other areas of the law, began establishing various checklists for courts to walk through when making decisions — lists of factors, or steps, or elements. In 1970, the Second Circuit boiled the contributory infringement decisions down into a pair of elements: “[O]ne who, [1] with knowledge of the infringing activity, [2] induces, causes or materially contributes to the infringing conduct of another, may be held liable as a ‘contributory’ infringer.”

A test this broadly worded is subject to varying interpretations. “Knowledge” could be specific or general, actual or constructive, detailed or predictive; “material contribution” could be anything that assists the infringer, anything that is necessary for the infringement, or an amount of assistance that surpasses some sort of importance threshold. (Post-Gershwin courts simply ignored the reference to “causing” infringement, and “inducement” was similarly overlooked until Grokster in 2005.)

As long as courts apply the test case-by-case, as a sort of rough guide, it doesn’t matter much whether a strict reading is applied in one case and a broad reading in another. Your Thanksgiving shopping list says “cranberries.” Is that whole cranberries, cranberry relish, or the cylindrical jelly? Use your best judgement; to each their own. But if the test is some sort of sacred text, it has to be read the same way every time. Once a court decides that “knowledge” requires certainty, it requires certainty in all instances. Once “material contribution” is defined in one case as simply some tool that is used in infringement, it means that in all cases.

This is the problem that courts in the 2000s started having applying the Gershwin test to modern fact patterns. A pair of cases, months apart, in the Ninth Circuit make the dilemma clear. In Perfect 10 v. Amazon, the court held that a search engine provided a “material contribution” to infringing websites by including links to them in search results: “There is no dispute that Google substantially assists websites to distribute their infringing copies to a worldwide market and assists a worldwide audience of users to access infringing materials.” But in Perfect 10 v. VISA (a second case brought by one-company copyright wrecking crew Perfect 10 magazine), a different panel on the same court held that processing credit card payments for infringing sites was not a material contribution. Payment processing just seems different, but the majority had difficulty articulating exactly how. The credit card companies “have no direct connection to that infringement,” they wrote:

[T]he services provided by the credit card companies do not help locate and are not used to distribute the infringing images. While Perfect 10 has alleged that Defendants make it easier for websites to profit from this infringing activity, the issue here is reproduction, alteration, display and distribution, which can occur without payment.

Money may make the world go round, but not all the way to contributory liability.

I’ll continue tomorrow with the fallout from the Supreme Court’s 2005 Grokster case, which didn’t clearly replace Gershwin and isn’t clearly consistent with it either.