In my first post, I discussed the emergence of the Gershwin test and how it’s run into trouble from a combination of rigid interpretation and novel fact patterns. In my second post, I argued that this problem was made worse with the Supreme Court’s Grokster decision, which cited Gershwin and referred to contributory infringement, but discussed only intentional inducement.
The Cox case brings the contributory liability question back before the Supreme Court for the first time since Grokster. That makes it an ideal opportunity for the Court to straighten out some of the confusion, but there is always the danger that a generalist, textualist court could instead make things worse. (See, e.g., Star Athletica v. Varsity Brands.) Doing my bit to try to avoid that result is part of the reason I spent the 100-plus hours to submit an amicus brief in this case, but not the entire reason. After all, the court gets dozens of amicus briefs in a case like this, so the marginal impact of an additional brief is near zero. (One oddity of the Supreme Court rules is that every brief, even a pro se amicus brief, is required to be filed by a “counsel of record.” So I filed the brief as counsel to myself.)
The other reason I bothered to write my brief is because I spotted a connection that I don’t think has been fully presented by anyone else. I’ve been pondering the relationship between indirect copyright liability and tort law for over a decade, so it captured my attention at the cert. stage in this case when both Cox and the Solicitor General relied heavily on Twitter v. Taamneh, decided in 2023. Taamneh had nothing to do with copyright law. Instead, the case involved claims against Twitter, YouTube, and Facebook under the Justice Against Sponsors of Terrorism Act (JASTA) for “knowingly providing substantial assistance” to persons engaged in international terrorism. The Taamneh complaint alleged that the platforms knew members of ISIS were using their services, but did nothing to remove them. The Supreme Court held that that was insufficient for liability under JASTA. Otherwise, the Act “would effectively hold any sort of communication provider liable for any sort of wrongdoing merely for knowing that the wrongdoers were using its services and failing to stop them,” a conclusion that would “run roughshod over the typical limits on tort liability.”
Aha, Cox and the Solicitor General said in their cert. briefs, that’s exactly like Cox! Sony Music and some amici, on the other hand, argued that Taamneh was decided under a completely different statute, and therefore of dubious applicability to a copyright infringement claim, particularly one with more compelling facts about the defendant’s involvement.
I don’t think either side of this debate has it quite right. Cox and its amici are correct that there’s a deep connection between civil aiding and abetting liability, the subject of a lengthy analysis in Taamneh, and contributory liability in copyright law. But that connection has to do with the legal doctrine and how it’s applied. Sony Music and its amici are correct that factually, this case is far different from Taamneh, in a way that justifies sending it to a jury — which it was, and the jury had all the tools it needed to decide the case under a civil aiding and abetting framework. I’m a bit ambivalent about the use of juries to decide complicated copyright policy questions, but the Supreme Court for the most part isn’t, and this case went to a well-informed and properly-instructed jury that decided against Cox.
First, the connection. It’s impossible to know exactly what the Second Circuit judges on the Gershwin panel in 1971 had in mind when they devised their test for contributory liability. All the opinion says, after a discussion of vicarious liability, is that “[s]imilarly, one who, with knowledge of the infringing activity, induces, causes or materially contributes to the infringing conduct of another, may be held liable as a ‘contributory’ infringer.” There’s a footnote after “materially contributes” that warns, quoting the Supreme Court’s decision in Fortnightly v. United Artists, that “mere quantitative contribution cannot be the proper test.” But there’s nothing indicating where the two-part test for contributory infringement came from. The only hint is a discussion of a district court opinion by Judge Edward Weinfeld — famously, a judge other judges went to with their most difficult legal questions — in which Judge Weinfeld referred to “common law concepts of tort liability” in determining whether an advertising agency, a broadcaster, and a shipping firm could be held liable for assisting a record piracy operation.
Gershwin is thus like most opinions written in the 1970s and earlier, in that it is extremely terse about the sources supporting its reasoning. But it’s not actually that hard to determine what “common law concepts” Judge Weinfeld and the Gershwin panel may have been referring to. The contributory infringement test in Gershwin closely resembles the formulation for civil aiding and abetting liability set forth in the first Restatement of Torts, published three decades prior to Gershwin in 1939. Section 876 provides that a person is liable “[f]or harm resulting to a third person from the tortious conduct of another,” if the person “(b) knows that the other’s conduct constitutes a breach of duty and gives substantial assistance or encouragement to the other so to conduct himself.”
In other words, civil aiding and abetting liability, as defined in the Restatement of Torts since 1939, requires knowledge plus substantial assistance. The knowledge element is identical to Gershwin’s knowledge element: knowledge that someone else’s conduct constitutes a breach of a duty to a third party. And “substantial assistance” is practically a synonym for “material contribution” — “material” meaning, as it does in Rule 56, “important” rather than “physical.”
And it is civil aiding and abetting liability that the Supreme Court explained at length in Twitter v. Taamneh. Several points in that discussion are notable. One is that the Taamneh Court emphasized repeatedly that civil aiding and abetting requires an overall determination of culpability that should not be hidebound to the letter of any particular list of factors or elements. For example, in passing JASTA, Congress referred to the D.C. Circuit’s decision in Halberstam v. Welch, which it described as “the leading case regarding Federal civil aiding and abetting and conspiracy liability.” But the Supreme Court chastised the Ninth Circuit in Taamneh for “hew[ing] tightly to the precise formulations that Halberstam used…. [A]ny approach that too rigidly focuses on Halberstam’s facts or its exact phraseology risks missing the mark.” Even knowledge and substantial assistance are not to be analyzed independently: “‘the knowledge and substantial assistance’ components ‘should be considered relative to one another’ as part of a single inquiry designed to capture conscious and culpable conduct.” The greater the defendant’s knowledge, the more acts might be considered “substantial assistance,” and the greater the assistance, the less detailed the knowledge needs to be.
Relying on the D.C. Circuit’s opinion in Halberstam — an opinion written by Judge Patricia Wald, joined by Judges Robert Bork and Antonin Scalia — the Taamneh court also stressed the value of the list of factors the Halberstam court used to determine whether assistance was substantial. Those factors, most of which date back to the First Restatement, include “(1) ‘the nature of the act assisted,’ (2) the ‘amount of assistance’ provided, (3) whether the defendant was ‘present at the time’ of the principal tort, (4) the defendant’s ‘relation to the tortious actor,’ (5) the ‘defendant’s state of mind,’ and (6) the ‘duration of the assistance’ given.” As with the two main elements, these factors should not be used as a mechanical checklist of “disparate, unrelated considerations without a common conceptual core,” but rather as a means to determine “the essence of aiding and abetting: participation in another’s wrongdoing that is both significant and culpable enough to justify attributing the principal wrongdoing to the aider and abettor.”
Firming up the connection between contributory copyright liability and civil aiding and abetting would have a number of benefits. First, it would establish that Gershwin’s two elements of knowledge and material contribution are not to be considered in isolation. Trivial amounts of knowledge, combined with routine acts of assistance, do not equal material contribution. Either more knowledge or more particularized assistance will be necessary. But by the same token, liability does not require the highest levels of both.
Second, following civil aiding and abetting law would clarify the required state of mind required for contributory liability. For the majority of courts, both in the tort context and in copyright cases pre-Grokster, liability follows from providing assistance with knowledge of an impending tort, but not necessarily an intent to further it. The Restatements have consistently referred to knowledge as the required scienter. As the reporters for one part of the Third Restatement of Torts have put it, “[i]t is well-established that knowledge, but not intent, is required for civil aiding and abetting liability.”
This contrasts with an argument presented by a number of my friends and colleagues at other schools, that only two forms of fault-based tort liability exist: intentional torts and negligence. Therefore contributory copyright liability requires proof of either intent or negligence. These scholars propose replacing the Gershwin test with a new formulation, tentatively named “negligent contributory infringement,” that would require a showing of failure to take reasonable precautions against foreseeable infringement that proximately caused that infringement.
But I think this argument gets off on the wrong foot. It’s clear that not all tort liability has to fit into the traditional division of primary torts into strict liability, intentional, or negligence. Just because aiding and abetting does not require intent (and is clearly not strict liability) does not mean that it’s based on negligence. There are sections of the Restatements devoted to negligent enabling of someone else’s tortious behavior, see Restatement (Second) of Torts §§ 302A, 302B; Section 876(b) on civil aiding and abetting is not one of them. That is why the Third Restatement Miscellaneous Provisions draft concludes that “[b]ecause aiding and abetting liability is based on knowledge rather than an intent to cause harm or merely objectively unreasonable conduct, aiding and abetting liability cannot be classified as either an intentional or negligence tort.” In any event, it’s unclear why as a conceptual matter it should not be possible to base secondary liability on knowing assistance of someone else’s tort.
That leads to the third benefit of connecting contributory liability to civil aiding and abetting: a set of factors to help courts determine when a secondary actor is providing a material contribution. This is the issue the Ninth Circuit struggled with in the Perfect 10 cases in 2007: why are search results a material contribution, but payment processing is not? The Halberstam factors help to answer that question. It’s not an issue, as Gershwin itself stated, of the sheer amount or necessity of the assistance. Rather, the Halberstam factors assess what sort of act the primary tortfeasor was engaged in — that is, how obvious the wrongdoing was; the quantity of assistance; whether the defendant was present at the time, meaning what sort of opportunity the defendant had to assess the situation and intervene; the nature of the relationship between the defendant and the primary tortfeasor, indicating the degree of control; the defendant’s state of mind; and how long assistance was given.
Cox has argued that Taamneh suggests a categorical rule against contributory liability for communications providers. But the nature of both the Halberstam factors and the interdependence of the overall test belie that assertion. While it is true, as Taamneh said, that communications providers do not “incur culpability merely for providing their services to the public writ large,” the words “merely” and “writ large” do a lot of work in that sentence. As the Taamneh Court also said, “[b]y their very nature, the concepts of aiding and abetting and substantial assistance do not lend themselves to crisp, bright-line distinctions.” Even routine services can become substantial assistance with enough knowledge of what the recipient of those services is up to.
The problem for Cox is that the facts concerning its knowledge and assistance are far more compelling than the allegations against social media platforms in the Taamneh case. In Taamneh, the plaintiffs’ allegations were that the platforms were aware that terrorist groups were using their sites. But their harm was caused when their relative was killed at a specific attack in Istanbul. There was no allegation that that attack had been planned on social media, and thus no way for the platform owners to have known of the planning. Instead, the plaintiffs alleged only that the platforms provided only general communications services to terrorist groups, who they knew to be out there somewhere among their users.
By contrast, the jury in the Cox case heard days of testimony and saw numerous documents concerning the notices Cox received from the anti-piracy firm hired by the record labels and music publishers, what sort of infringement was alleged, Cox’s response to the notices internally and externally, Cox’s relationship with its subscribers, the harms of interrupting service and the reasons why Cox might have not suspended or terminated subscribers, and the length of time this process continued. In other words, even though Cox was not tried as a civil aiding and abetting case, the jury heard testimony and argument on every aspect of such a claim, including the Halberstam factors, meaning that even if the Court should decide to incorporate civil aiding and abetting into contributory liability doctrine, there’s nothing to send back to the jury.
The jury even heard extensive testimony on the specificity and reliability, or lack thereof, of Cox’s knowledge of infringing activity, which had already been decided by the court on summary judgement and thus was technically not before the jury. As the Court noted in Taamneh, knowledge and assistance are intertwined, and the jury had to consider Cox’s knowledge as part of its determination of culpability. And after weighing the evidence both in favor of Cox and against it, the jury reached a conclusion on that culpability.
I once heard a lawyer explain to her clients, “We are excellent attorneys. And we will make the best legal arguments possible for you. But we just have bad facts.” Cox went to trial with bad facts, and it’s not clear basing contributory infringement liability on the test for civil aiding and abetting will change that.