Cox Communications v. Sony Music Entertainment

Thank you, Mr. Chief Justice, and may it please the Court: The Fourth Circuit held that a provider of basic communications infrastructure to millions of homes and businesses can be held liable because it did not kick enough accused infringers off the Internet. No notion of tort or copyright law ever conceived can support that theory. This Court explicitly rejected the theory in Twitter, where it said, "Plaintiffs have identified no duty that would require communication-providing services to terminate customers after discovering that the customers were using the service for illegal ends." This Court said in Grokster that liability cannot be predicated on "mere failure to take affirmative steps to prevent infringement." Reaffirming those basic principles resolves this case. No case has suggested that knowledge alone can create the necessary culpability to -- to find someone liable for infringement. The case -- the consequences of Plaintiffs' position are cataclysmic. There is no sure-fire way for an ISP to avoid liability, and the only way it can is to cut off the Internet not just for the accused infringer but for anyone else who happens to use the same connection. That could be entire towns, universities, or hospitals. Turning Internet providers into Internet police for all torts perpetrated on the Internet will wreak havoc with the essential medium through which modern public engages in commerce and speech. This Court should reverse. I welcome the Court's questions.

Well, Your Honor, the -- the question of how far one goes with Twitter, when you think about the different --

Oh. So -- so Twitter held several things that are -- that are fully applicable here. The first is that contributory liability requires malfeasance with the purpose of fostering the bad act, that there's no liability for passive non-feasance, and there's -- there's no liability for sales to the general public on the same terms regardless of any eventual use. Twitter is -- we -- we are an a fortiori case of Twitter. Twitter at least had those communications on its own server. They were visible to Twitter. What our -- what our customers do is invisible to us in real time. Now knowing that a particular account -- account will infringe does not make us an accomplice, which is what's required for contributory liability. Continuing to supply the Internet is just not the same as joining with the tortfeasor in "mind and hand" in the infringement as in something that the ISP wants to bring about.

Your Honor, the basis of contributory liability for copyright, as this Court recognized in Sony, is just sort of adopting the common law in a world in which the Copyright Act itself does not prescribe secondary --

Well, Your Honor, it must be, I would say, the narrowest version of common law that is available precisely because Congress did not speak directly to it. And so I would say Halberstam merely articulates a way of getting at the ultimate principles that the Copyright Act -- excuse me, that contributory liability is directed at, but Twitter and Smith & Wesson apply it to the exact sorts of circumstances here, which is where you've got an arm's-length seller selling to millions and millions of people --

Well, so, Your Honor, it's -- it's a step further in one respect and it's a step in the other direction on the other end of the spectrum. So, on one end of the causation spectrum, we can't see in real time what people are doing. That, as I said, distinguishes us from Twitter and makes us more remote. You are right that on the other end of the spectrum, there's a slightly greater nexus, but I hasten to add it is still a limited nexus. When we get pinged about a regional ISP with tens of thousands of users, we know that somewhere someone in that community has infringed.

Your Honor, I beg to differ. There is literally not a single place in this record where a specific individual was identified. If you -- let's take the smallest unit, a household. You still don't know who the individual is. I also beg to differ on the first half of what you said, Your Honor, that there's no evidence that what I've described has happened here. Page of our opening brief gives you the -- the nature of the highest recidivist infringers. They are regional ISPs, universities, nine hotels, and so forth. Those are the entities that are most likely to be cut off first because those are the ones that accrue the greatest number of --

So, Your Honor, a couple of things. First, on the -- on the first point you made, the DMCA does purport to require termination, and it is the failure to terminate --

No, no, I -- I'm talking about the DMCA, the statute itself --

-- the safe harbor. And that was the theory of liability, that we did not terminate people enough. Plaintiffs in their brief to this Court for the first time have made the assertion that you will find nowhere in the court of appeals brief that we did nothing. The notion that Cox did nothing is absurd. I will mention just three facts that are undisputed. First, Cox invested its own resources to create the first-of-its-kind anti-infringement program. There was no precedent for that. Second, under that program, Cox sent out hundreds of warnings a day. To your point, Your Honor, that we didn't work with universities, we most certainly did. The first several steps, the steps, are all about contacting them, cutting them off, that is, suspending their accounts, which we did 67 times -- 67,000 times in the course of this period. That's thousands every month -- month. And, third, the program stopped infringement by 98 percent of the people who were accused of infringement. That is not nothing, Your Honor.

Well, so, Your Honor, you're asking what's -- what is the problem with the simple measures test?

Oh. Well, so --

We most certain -- yes, Your Honor. Sorry. I did not mean to --

So what is wrong with that test is that it does not comply with what this Court has said and the common law has been saying for a hundred years.

I'm -- I'm not complaining about the practical impact on us. We easily meet the simple measures test. And I would underscore that Plaintiffs have not asked this Court to apply the simple measures test and have not argued that we could not meet the simple measures test.

Your Honor, for the simple reason that Cox is a good corporate citizen that cares a lot about what happens on its system. We do all sorts of things that the law does not require us to do.

That is correct, but I want to underscore my -- my answer to --

So -- but, Your Honor, my answer is no. And -- and I agree with you that the way one conducts this analysis is not to say, oh, purpose is the test. The way one conducts the analysis is to assess --

No, I don't think the government is saying that purpose is the standard, but we -- what the government --

What -- what the government and we both contend is that there are acts that can be -- from which one can impute intent, and in the context of a seller, in particular, a seller of communications technology, this Court has held that you -- you -- that -- that unless it is a technology that is incapable of multiple uses, that is, non-infringing uses, you don't infer intent from what is an --

No, you -- you don't have to go that far to recognize that -- that both the jury instructions here and, in particular, the Ninth Circuit's holding that the instructions as applied to this set of facts could -- could sustain liability.

There's nothing wrong with the syllogism. Just one quibble. We're not challenging the jury instructions. We're making a JMOL argument. But, yes, everything else you said I completely agree with, that what the -- what the Fourth Circuit did was to say that continuing to provide Internet service with the knowledge that someone on that account is likely to infringe again, that that is the culpable act. And that is just absolutely not correct. This Court rejected that exact theory in Twitter. I would remind the Court to the questions asked earlier about whether Twitter goes that far. There was knowledge in Twitter of actual accounts. I understand it was not knowledge of specific acts of terrorism, but -- but -- but knowledge was very much a part of that case.

Your Honor, the position we're taking is not that an ISP can never be culpable. But an ISP can be culpable only if it engages either in clear expression, such as inducement, or in affirmative acts that align itself --

That is correct because that is simply another way of packaging the non-feasance.

That's correct, Your Honor. That's a difficult hypothetical that pushes the envelope.

Yeah, understood. Let me just underscore the facts you recited are -- are facts that would make it totally plausible and, in fact, would invite the I -- the Plaintiffs here to sue that infringer directly. They have recourse in a situation like that.

No, Your Honor. The -- the reason that it wouldn't be enough for Cox in this case is exactly the reason that it wouldn't be enough under the common law. The common law also requires a culpable act, not the equivocal act of selling to everyone on equal terms. Smith & Wesson says that, and Twitter says that too.

That is an equivocal act that does not align oneself, that does not align the seller with the purpose of promoting infringement.

Your Honor, hypothetically, I can imagine that being true, but the facts of this case are percent of traffic to the other ISPs was infringing. We were two-thirds of that. So the -- the notion that we had a -- a particularly impressive track record of reducing infringement did -- did not deter us from -- excuse me -- of reducing infringement did not deter us from having a program that worked to the point of 98 percent success.

We -- we have a financial incentive, like any business does, to keep our customers. But the Fourth Circuit found -- and this was a basis for rejecting vicarious liability -- that we did not have a financial incentive to increase infringement.

No, Your Honor, because there could be all sorts of reasons why an ISP does not want to cut off a customer. And it's not just a customer who, in your hypothetical, confesses that he's going to continue infringing. I think that -- another way of answering the question is think about the flip side. In order to capture that fringe case, we would have to create a world in which ISPs are going after everyone who infringes, including not just that one individual user of which we have no evidence in this record that we ever knew about that one person but the regional ISPs and the hospitals and the universities, because that is a world in which, at $150,000 per pop, the ISP has a huge financial incentive to cut off the accused infringer at the word "go." I will underscore that -- that in that table on page 11, of the 49 infringers who had more than a hundred notices, only one was a single-unit home. Everything else were the sorts of businesses and multi-unit dwellings that I described.

Well, so two --

Well, so -- so a couple of answers, Your Honor.

So -- so a couple of answers. I -- I certainly didn't concede that there would be liability. I said, of the top recidivists, there was only one single-family home. We still don't know the identity of the --

Excuse me?

Well, because, Your Honor, it needs to be an act that unequivocally demonstrates --

-- a purpose.

Your Honor, I'm not just talking about the Sony case. Grokster reiterated the same point, Twitter. If the rule was that there -- that -- that the rule is different if there's an ongoing relationship, then Twitter would have come out differently. By the way, so -- Direct Sales underscores that point too, and that was 80 years ago. That is part of the --

No, Your Honor.

This Court explained in Smith & Wesson why it's different, and that was -- and it said explicitly it was not just that there were -- that there was an ongoing relationship. There were on --

Well, can I just answer the question, though? I'm not citing Swith & Mess -- Smith & Wesson in answer to your question. I'm citing Smith & Wesson for its analysis of Direct Sales. And what the Court said in Smith & Wesson about Direct Sales was ongoing wasn't enough. It was the stimulation of further sales and -- and treating that buyer differently from others, that was the key that created culpability.

That's correct. But I was making the point earlier about the corporate citizen in response to the question why would an ISP continue to provide anti-infringement programs if they don't have to, and I was simply saying --

It is irrelevant, both the invective from Plaintiffs and my protestation --

Well, so, Your Honor, Congress can --

Well --

Right. So -- so Congress can adopt a safe harbor for all sorts of reasons. Here, Congress did it to assure service providers at a point in time at which the law was really unknown.

Twenty-seven years later, the safe harbor provision, by light of this Court's ruling that came a generation after the --

Correct. Yes, I agree.

That is correct, with one caveat. It can be shown when it is -- when -- when the only use of the -- of the device is an infringing use.

Correct.

So there are two things that are good enough. One is inducement, which are mainly -- generally words that are directed at encouraging infringement. The other is affirmative conduct directed at -- at fostering infringement. For example, if an ISP were to provide an anti-infringement detection buster, you know, something that hides the ISP address from the MarkMonitors of the world that detect infringement, that would be an affirmative act that's built into its device that aligns itself as someone who desires for the infringement to occur. Another possibility is if the ISP creates a service that is uniquely designed to help individuals take streaming services, the songs off of streaming services, download them, and swap them. Those are two -- and they're classic examples of affirmative conduct that amounts to material contribution.

Oh, a -- here, let me help you infringe. We have an infringement hotline.

Sure. Or even an employee who will instruct people how to go about getting BitTorrent on their computers.

So I think that that's a hard case. I don't know what -- how Twitter would come out on that. But everything Twitter said about affirmative conduct and malfeasance versus non-feasance, I would say, would drive the -- the conclusion the same way that Twitter ultimately drove it. Now I can imagine a court -- this Court someday carving out an exception for life and limb or imminent danger of physical harm, but --

That's correct. I mean, this Court said in Twitter what I said right at the outset of my argument, that Plaintiffs have identified no duty that would require communication-providing services to terminate customers, terminate them after discovering that the customers were using the service for illicit ends. I'm just reading what this Court said in Twitter.

I -- I would think so, Your Honor. This Court has always treated those two lines of cases interchangeably. When this Court asks what the common law is of copyright contributory liability, it is looking to, back -- you know, back to Kalem, accomplice liability. So what this Court says here will apply to other contexts, except if there's a statute that changes the common law. And vice versa, what this Court said in Twitter and in Smith & Wesson should apply with full force to this version of common law liability.

So, Your Honor, it's called the abuse team and so it tracks all sorts of abuses. It got 5.7 million notices in the particular period. It was reacting not just to copyright but to the point I was beginning to make earlier. It reacts to phishing, it reacts to fraud claims, it reacts to hacking, none of which there is a legal duty to act.

No, absolutely not, Your Honor. We embrace the full breadth of contributory liability that includes inducement and includes material contribution. And I gave Justice Kavanaugh a few examples of thing --

How it would work in -- in the ISP context. I'll -- I'll hasten to add, though, the fact that there are not a lot of examples is a function of Plaintiffs' decision to sue the one technology that is least likely to fit a fact pattern of material contribution. A -- an --

Oh, it's definitely narrower than what the Respondents will say because Respondents have never purported to identify an affirmative act to foster infringement. Their argument is knowledge alone is what -- what satisfies --

That is correct. We are providing the Internet service and declining to terminate. That's what we were held liable --

So -- so contracts, yes, but there was an ongoing relationship in Twitter with all of the customers --

We know the IP address. If it's a regional ISP --

-- there are 10,000 possible homes or businesses who could be infringing, and, as I was saying earlier, that's the first that will have to get cut off under this liability scheme.

So -- so two answers, Justice Jackson. The first is, just to continue the answer I was giving earlier, Congress had no clue what the liability would end up being against four different types of service providers for -- back then, service providers were confronting all sorts of theories of liability. It included direct liability because you have to copy --

Right. So -- so a couple of answers to that. First -- I will get to the legislative history in a moment, but, first and foremost, what Congress said was in 512(l) that the failure to satisfy a safe harbor "shall not adversely bear on liability." And it punctuated this point in the Senate report. It was "leaving current law in its evolving state" because it had no idea and most certainly couldn't agree upon --

Your Honor, Congress had no idea what the liability would allow. There was no case at that point providing for liability for ISPs, that is, holding them liable ultimately. I will also underscore there was no such thing as a conduit ISP. That was a creature of 512(a). The AOLs of the world were hosting content. So Congress -- you -- you -- one cannot impute to Congress any view and certainly not any unified view about what the liability rules would be in the absence of the DMCA.

Thank you, Mr. Chief Justice, and may it please the Court: In Grokster, this Court held that one who distributes a device with the object of promoting its use to infringe copyright can be held liable for its users' infringement. In Twitter and Smith & Wesson, the Court emphasized that to be liable for aiding and abetting, a person must participate in the primary violation as in something that he wishes to bring about and seeks by his -- seek by his action to make it succeed. Thus, both in copyright law and more generally, this form of secondary liability is reserved for persons who act for the purpose of facilitating violations of law. Because Cox simply provided the same generic Internet services to infringers and non-infringers alike, there is no basis for inferring such a purpose here. I welcome the Court's questions.

I think probably the first place you would look is general principles of aiding-and-abetting law, which are common law principles. But, if the Court thinks maybe intellectual property cases are different, it could look to Congress's codification of secondary liable -- liability principles in the Patent Act. And it's very clear that in patent cases, the Patent Act doesn't provide for secondary liability in circumstances like these. The Patent Act in 35 U.S.C. 271(b) says whoever actively induces infringement of a patent is liable. And Section 271(c) says anyone who sells a device that is specially suited to patent infringement, not capable of substantial non-infringing uses, with knowledge of its special character, is liable. There's nothing in the Patent Act that says, if you sell a multi-use device to a person that you know is going to use it for infringement, you can be liable on that basis. So I think it would be unusual to adopt a special rule of copyright law and ignore the rule of patent law that Congress has actually adopted in the U.S. Code.

I think -- yes. I mean, I think the patent law, as it's currently codified, incorporates the two bases for secondary liability that the Court adopted in Grokster and Stony -- and Sony as principles of copyright law. That is, in Grokster, it said the -- the -- the law that one who induces infringement can be held liable was a common law principle before it was codified in 1952, and we adopt it here as a principle of copyright law. Similarly, in Sony, said we adopt the staple article of commerce doctrine from patent law as a principle of copyright law, that ordinarily, if you sell a staple article of commerce, one capable of both infringing and non-infringing uses, you won't be liable for the misuse made by individual customers. And I -- to -- to -- I'm sorry.

I was going to say, to follow up on Twitter, the -- the two things that we think -- we agree that the suit in Twitter had problems that -- in addition to the ones that we're noting here; that is, even if the platforms in Twitter had highlighted content celebrating ISIS, they could have been liable for aiding and abetting a particular attack only if a causal connection could be shown between those videos and the attack, and there was no such proof. But I think two aspects of Twitter are deeply relevant. The first is five different times the Court in Twitter quoted in whole or in part the standard that I just read that a person who aids and abet -- who is liable for aiding and abetting must be shown to have participated in the wrongdoing as something he wishes to succeed. And that standard can't be satisfied here. The second thing the Court in Twitter emphasized time after time was that, at worst, the platforms had simply treated ISIS the same as they had treated other users, the --

It -- it --

-- it wouldn't be secondarily liable. That is, I think lurking in Twitter was the idea -- and the companion case, Google versus Gonzalez -- was the idea that perhaps a platform like YouTube or Twitter, a platform on which potentially infringing or otherwise unlawful content actually appears, the platform could be held directly liable on the theory that it has -- it was speaking the words as well as assisting the -- the third-party content provider in doing so. And this is a little far down in the weeds, but, in the Communications Decency Act of 230 -- of 1996, 230(c)(1) said generally that platforms that host third-party content will not be treated as the publisher or speaker of that third-party content. But there's --

For purposes of aiding and abetting, the answer is the same. And -- and, obviously --

-- direct copyright infringement is a strict liability offense. That is, if you think of the platform as the person who's actually speaking or publicly performing the infringing video, then it doesn't matter whether the platform wants to infringe. It can be held liable on that basis. But, because Cox just provides the infrastructure, there's no plausible theory of direct liability.

Right.

Yes.

I think you can still that -- do that with respect to what I would call targeted assistance, that is, if you are providing assistance to a particular person and you're not providing it to anyone else and you know that person is using that assistance to commit a violation of law.

Yes. And -- and it -- yes. Exactly.

Those are the only ones that are codified in the Patent Act.

Right.

I -- I -- I -- we're not asking the Court to disturb the -- the existing body of copyright law, except with respect to ISPs. That is, in Sony, the -- the Court noted with seeming approval that there were various lower court cases involving things like dance halls where people were held secondarily liable --

Yes, and -- and particularly with what I would regard as kind of at the opposite extreme from non-targeted assistance. That is, in cases like Grokster and Twitter and in this case, we're looking at a pretty narrow subset of aiding and abetting. We're asking when can a person who provides a good or service to the general public be liable for misuse of that good or service by some member of the public, and a lot of --

I think you could absolutely do that. And the flaw in the court of appeals decision was that it acknowledged the purpose language, but it said we typically infer a purpose to cause a particular event from knowledge that the event is substantially certain to occur. And I think there are bodies of law where we care more about knowledge than purpose or objective. But there are others where we don't. And, for instance, in equal juris -- protection jurisprudence, there's a fundamental difference between a state doing something because it will produce a racially disparate impact and doing something for race -- race-neutral reasons even though it will produce a race -- racially disparate impact.

And --

I mean, they --

I -- I -- I -- I agree with that. And the ISPs might not bother. I -- I would push back a little bit against the assumption that good -- good corporate citizenship would necessarily mean terminating repeat infringers. Imagine a case in which a person -- the individual was sued as a direct infringer and a jury found that on multiple occasions this person had used the Internet to commit direct infringement. Clearly, the Court could award damages and, under the Copyright Act, the Court would have authority to fashion an injunction that was reasonably designed to prevent further infringement. But could the district court enjoin the person from ever using the Internet again? I don't think so. I think the -- the general rule that equitable relief is supposed to be tailored to the wrongdoing --

I -- I --

I -- I -- I would agree that not much economic incentive would be left. I'm simply questioning whether that's a bad thing. That is, if we don't think a district court could enjoin a repeat infringer from again using the Internet, I don't --

They -- they provided the safe harbor, but they also provided kind of the takedown notice provisions of the -- the DMCA, and nothing -- no notices like the ones that were sent here are contemplated by the DMCA. The DMCA does contemplate notices in which different kinds of Internet company may be informed there are particular infringing works on your platform and we want you to take those down. And if the Internet companies do that, the result is a much more targeted approach. It's that you get rid of the infringing materials, but the rest of the platform remains intact and people can use it. The -- the approach of terminating all access to the Internet based on infringement, it seems extremely overbroad given the centrality of the Internet to modern life and given the First Amendment.

Well, the -- the advantage, I guess, is the same one that the Court perceived in Twitter, whereas the Court held that the case should have been dismissed on the pleadings. And the idea was a suit in which all you could show was that the defendant had provided kind of generic services to the general public and that some people had misused it and the -- the defendant had failed to make that stop, that wasn't enough even to get past 12(b)(6). And so I think there -- there would be an advantage to making clear that in the future, not only should the -- the ISPs win at the end of the day, but they shouldn't be subject to the burdens of litigation if -- if that's all you have.

I -- I mean, I guess the -- the thing I would emphasize, and I'm sorry, I don't remember the -- the precise deal -- details of his answer, are I would emphasize the -- the differences between targeted assistance and provision of general purpose technology. That is, there are an infinite --

There -- there are an infinite number of situations in which somebody may provide assistance to one person that he isn't providing to anyone else with --

Such as, in the dance hall cases, the -- the proprietor of the dance hall was hiring a band to play and the band was --

I mean, the -- I don't think there would be an ISP equivalent unless, as Mr. Rosenkranz suggested, the company -- the -- Cox provided some special service that would only be useful for infringers. In that case, you could say that this is targeted at -- at infringement. The point about the -- the targeting is often, when you do something for one person that you don't do for anyone else and you know that person will use the assistance to commit a legal violation, it's reasonable to infer that's why you provided the assistance. But, when you are providing Internet service to every member of the public who will pay the fee, infringers and non-infringers alike, there's no basis for that --

I -- I guess I'd make two different points about that. The first is, as Cox's reply brief points out, if you look back at the trial court opinion in Henry, the court's opinion recites that the ink that was being sold was a specially suited ink, specially developed for a mimeograph of this type. And the trial court opinion also reflects that the defendant went to the offices of the purchaser and told her, I'll sell you this ink for your mimeograph, but pour it into an old jar that was provided to you by the patent holder and throw my jar away. And so the word "expectation" originally came from the -- the trial court's opinion. That was the certified facts. And so I think Henry on its facts was a good case for secondary liability. The other thing I would say is, even if you read Henry as standing for a broader proposition, namely, that if you sell a multipurpose product to a particular customer that you know will use it to infringe, you can be held liable, that's been superseded by the Patent Act, in which Congress codified other rules of contributory infringement in patent case but not that one.

Right.

I -- I guess, in part, I'll -- I'll want to hear Mr. Clement's presentation because I would want to know does he think there is a plausible basis on which his client might still seek to prevail under a purpose standard when purpose is understood really intent to bring the result about and desire that it succeed. If Mr. Clement thinks we -- on the existing record we might be able to show that even in this case, that would weigh in favor of sending it back.

I -- I would just make it clear that you are limiting your holding to situations in which the plaintiff seeks to impose liability for selling a generic product to a mass audience, and in that circumstance, we're comfortable saying these are -- these are basically the only two ways. But we would not want the Court to foreclose the possibility that purpose could be shown in a variety of ways when you're providing what I referred to as targeted assistance.

I -- I -- I think Mr. Clement would probably say that the DMCA reflects a special focus on Internet companies and that it is a justification for inferring potential liability in that scenario and there's no DMCA equivalent for electric companies.

Thank you.

Mr. Chief Justice, and may it please the Court: This Court's cases recognize that liability for copyright infringement is not limited to direct infringers but extends to those who induce, cause, or materially contribute to the infringement of others. And a classic form of material contribution is to provide the means of infringement to a specific known infringer, knowing that infringement is substantially certain to follow. That combination of knowledge of a specific consumer and an ongoing relationship is critical to distinguish culpable conduct from simply engaging in a one-and-done sale of an item that can be used in a way to infringe but is generally used lawfully. Now, on this record, there -- it is beyond dispute that Cox provided the service to known infringers with substantial knowledge that what they themselves called habitual abusers would continue to infringe. That reality, along with a record chockful of Cox's admissions that it held the copyright laws and the DMCA in contempt, is what requires Cox to insist on the extreme position that they can continue to provide service to habitual abusers in perpetuity without consequences. That rule has nothing to recommend it and was admitted today would render the DMCA and the cooperation it is intended to foster a dead letter. Why bother with a safe harbor? Why limit lie -- why worry about a limitation on liability, which is the express text of the DMCA, if there's no liability to limit? Why bother cooperating with copyright holders? Why bother having a reasonable and appropriate system for taking down repeat infringers if you're allowed to behave entirely unreasonably? So, in all of this, you see that the position that's being advocated by Cox is a product of the record in this case. If Cox is right on the law, then Cox could take tens of thousands of copyright notices and throw them in the trash, and they could have its employees say "F the DMCA." That is, in fact, what the record says, which is why they're asking you for an extreme rule. I welcome the Court's questions.

So what would limit my approach is I would say that the provider of the service has to know that specified customers are substantially certain to infringe, and that is, to my understanding, the standard that's long prevailed in the trademark context under the Inwood case, where this Court, based on a lower court opinion by Judge Friendly that was based in part by a district court -- earlier district court opinion by Judge Wyzanski, which was based on the common law, says that you can have liability under the trademark laws if you either induce or you provide the means of infringement to a specific known infringer. So that is the standard. And I think, if you limit it to knowledge of specific known infringers and you require -- and not purpose but intent that requires you to know that providing the service to that customer will make infringement substantially certain, I think that provides a strict limit and plenty of guidance for the lower courts.

Well, Grokster wouldn't get me very far in this case because this Court in Grokster rested its holding on inducement. But, if you look at the briefs in Grokster, if you look at the oral argument and what the lawyer for Grokster himself told this Court, that was all because --

It was not. It was Mr. Taranto, who's a fine federal judge now.

And what -- what Mr. Taranto told this Court was that there -- that -- that the norm of contributory liability was actually material contribution. But that wasn't at issue in the Grokster case because the Grokster technology was structured so that Grokster didn't know about the infringement of any particular customer. So that's why inducement was where this Court essentially had to go in the Grokster case, but I don't think it meant to eliminate what it was told by all the parties, including the United States, and I did represent them, that the -- that material contribution was an important part of secondary copyright law.

Yes, it has intent.

It -- it -- you have to show -- that's where the substantially certain comes. You have to know that the person you're providing the service to is substantially certain to infringe.

So it's -- what I'm offering you is not just knowledge. It is intent, and it is intent under the common law. And if you look at the Restatement that was the governing Restatement at the time that both the '76 Act was passed and the DMCA was passed, it's Section 8A of the Restatement (Second) of Torts, and it's using the definition of intent for purposes of intentional tort, and all these aiding-and-abetting torts at common law are intentional torts. And it says there's two ways to show intent. One is what you would think of as purpose, that you actually have the design of carrying out the primary wrongdoer's purpose. But the second thing and is -- equally qualifies and is equally intent is when you provide or do an act knowing that certain results are substantially certain. Under the common law, you --

-- are charged with intent.

Well, I mean, they would also be inconsistent with Halberstam. Halberstam does not have a purposeful intent. In fact, if you look at the three factors of Halberstam, which we -- I think Congress at least took to be the platonic statement of the common law, I mean, it talks about knowledge. It doesn't talk about purpose at all. And, of course, it would have to. Linda Hamilton didn't have the purpose of killing Halberstam. She probably didn't even have the specific purpose of facilitating the sort of night ventures out by her live-in sort of paramour. So what -- what counts at the common law and is always counted at the common law is intent, and intent can be showed either by purpose and design, and so I'm -- I'm sure that the Smith & Wesson decision was carefully worded to capture that aspect of it. But I don't think there was any occasion there to jettison essentially half of the common law standard, which is this intent where you are charged with the sort of -- those consequences that are substantially certain to follow from your actions.

I would certainly hope not. Now it --

Go ahead.

I -- I'd be delighted to. And the copyright law, if you go all the way back to the Harper case in 1886, has this notion of material contribution. And material contribution is an absolute critical aspect of copyright secondary liability as it's developed over the years. And I think it's important to understand that, yeah, you're right, you know, Grokster, as I tried to explain, didn't focus on material contribution, but that's because the knowledge of specific infringing works was missing there essentially by design of the -- the product. Now Sony I think actually does talk a fair amount about material contribution, and what it talks about is -- it says a couple of things. In the text around Footnote 18, it makes clear that one of the things that made that body of case law inapplicable, and it talks about the dance hall cases and a whole bunch of common law cases, those aren't applicable there because there's no ongoing relationship. And it sort of -- and -- and so, in that context where it's a one-and-done sale, then you really are forced to rely on the patent context of -- of -- of -- of a staple article of commerce that you just launch in and you have no way of -- of -- of -- of ascertaining any knowledge. I think, though, it's very telling that in Footnote Judge -- Judge -- Justice Stevens refers to the trademark law. Now he says the trademark law is not as apposite there as the patent law, but that's because trademark law is actually narrower in most respects, but he specifically cites this Court's decision in Inwood Labs, and Inwood Labs could not be clearer that the standard is inducement or material contribution as defined exactly as I'm defining it, which is to provide the means of infringement to a customer you know is going to use it to infringe.

And I'll give you substantially certain on top of that.

So I -- I think the way that my position works most readily is that the -- the safe harbor, I think, provides plenty of room for ISPs to handle things like multi-user accounts, fraternities, and the like differently, and you can have a different policy under those. The way this case was litigated below, my friends on the other side, I mean, you know, the -- that chart on page that was a demonstrative that wasn't even introduced into evidence, they didn't make a big pitch that, like, our liability should be limited to just those customers. Part of the reason is because what the record does reflect in things that are in evidence, transcript -- trial transcript pages 810 to 811 shows that 95 percent of the customers covered by the 57,000 sort of customers that are part of the -- the universe that makes up the 10,000 infringement, 95 percent of those are residential customers. Only percent are business customers. So I think there's a reason they didn't try to limit their liability to the percent and they -- they took a position that was essentially we stand or fall together.

The -- the university is supposed to -- under those circumstances, the ISP is supposed to sort of have a conversation with the -- with the university. Now the ISP's policy to the university says you can't have -- you can't use this service or allow your service to be used for copyright infringement. So that's --

Well, look, I'm not sure that -- this record certainly doesn't support the notion that there are universities that have sort of undifferentiated service to 70,000 students or whatever the hypo is. I don't actually think that's how it works in practice.

Well, the way it works in practice is with, let's say -- let me -- let me take something that I know a little bit better like a hotel. And so, like, a hotel has lots of guests. So the hotel is provided Internet service and the hotel then can do things starting with terms of use, but a lot of hotels actually don't provide their guests -- at least in a normal way don't provide their guests with services at a speed that are sufficient to do peer-to-peer downloading precisely because they don't want to be in the position of having guests that are staying there largely so they can sort of upload and download copyrighted works. And if a particular hotel wants to be, you know, the --

I don't think it would be the end of the world if universities provided service at a speed that was sufficient for most other purposes but didn't allow the students to take full advantage of BitTorrent. I could live in that world. But, in all events, this isn't a case that's just about universities. We've never sued the universities. We've sued Cox. We've sued Grande, and I think it's worth, in thinking about the consequences of this case, I think it's worth taking a quick look at the brief in opposition in the Grande case, which is being held for this case because that's a case where the ISP did exactly what my friend's position would incentivize --

-- which is they did nothing with the notices.

I -- I don't think that we could. We certainly couldn't readily. And, again, I think that there's going to be a -- I mean, first of all, you know, the hospitals are in sort of a different position. They're kind of like more the intermediary where, you know -- and -- and there's no safe harbor that's specific to hospitals and universities. I mean, you know, Congress was very focused in 1998 on the role of ISPs in all of this and they wanted to create an incentive for the ISPs to adopt reasonable measures. And I think that would certainly accommodate measures that treat multi-user addresses quite differently from residential customers.

We would -- we would be without scalable functional recourse. And if you look at the Seventh Circuit's Aimster decision, like even back then, Judge Posner had a nice phrase for what direct infringement is, which is it's a teaspoon solution to an ocean problem. So, if my clients are limited to direct infringement actions, they are in very, very dire straits. But it's worse than that because the key thing about the safe harbor is the safe harbor is not only what gives the -- the ISPs an incentive to behave responsibly. It's also what gives them an incentive to come to the table and have negotiations with the content community. And if you look at the -- the Motion Pictures Association brief, they talk about this at length. And if you look at the report of the copyright office, they put together a 200-page report on 512 and the safe harbors, and what they talk about is how important it is for essentially both sides to have skin in the game.

So, if we win and they do that, then they're not going to be able to take advantage of the safe harbor. I cert --

I -- I think there's a concept in the law called willful blindness, Your Honor, and I think willful blindness would satisfy the common law standard for aiding and abetting. And so, if somebody can't -- you know, if -- if you have, like, the hammer of the month club and you keep on giving a hammer to somebody who is, you know --

I do.

Well, I -- I think it shouldn't for a couple of reasons. First of all, I don't think you're dealing with an implied cause of action. I think you are dealing with an express cause of action in -- in the Copyright Act, and the question is, to what extent does it reach sort of secondary infringement? And I think, at this point, it's too late to say that Congress hasn't endorsed it. Let me point to three things if I could. One is the addition in 1976 to -- in -- in the basic provision of rights to copyright holders, Congress added the words "to authorize." So it's no longer you just have the exclusive rights to do certain things, like copy and distribute, but you now have the exclusive right to authorize it. And for those that want to take a peek at a Senate report, that -- those words were added specifically to capture secondary liability. So Congress starting in '76 expressly adopted it. But then the DMCA, for reasons that we've already talked about, I don't think can be reasonably understood except against the backdrop of secondary liability. And Congress put specific weight -- and, again, this is the Senate report -- but it looked at three contemporaneous cases. There was the Playboy case. I think the case that's most instructive is the Netcom case, which was right -- it was a '95 case, and that was a case that said, well, for somebody who's like an ISP -- I'm sure my friend will say it's not exactly like the ISP today -- but somebody who's very much like an ISP, there was no vicarious liability because they didn't make money off of the infringing, but there was material contribution liability. So that's the evolving common law that Congress is passing. And so I think Congress embodies that in the DMCA safe harbor, but just as a cherry on the sundae, in the same DMCA provision that I only came across over the weekend but still is in there, U.S.C. 1201(c)(2) is another provision that sort of says that the anti-circumvention provisions are not designed to change the rules of vicarious and contributory liability. So I think there are multiple textual acknowledgments of vicarious and -- vicarious and contributory infringement in the code that distinguish this from your sort of classic --

-- implied-rights-of- action case.

Well, I think it would be a cautionary tale to not go beyond the common law.

Well, but that -- that was --

I don't -- I mean, you know, look, Central Bank of Denver, you already have, like, a made-up cause of action, and the question is, do we make up a bow on the made-up cause of action? I think that's different from what you have in a case like this.

No. I really do think here, given all the text that you have and the Court's precedents, which aren't nothing, that it's too late to sort of say this is all like an implied cause of action.

Look, it's always good to have caution.

But I would look -- I would look not just to the patent law; I would look to the trademark law, and I would take the trouble --

-- I would take the trouble to trace it back --

-- to the common law.

I don't think so, but --

I -- I --

Well, I think you and I just read the Fourth Circuit case differently. I read the Fourth Circuit opinion by Judge Rushing to specifically say the substantial certainty test is what's being applied and is what makes it --

Well, you've got a brief from some other intellectual property lawyers who --

-- you know, professors who say that was exactly right.

Oh, oh.

No, but I actually think it is, but it -- it's a little bit opaque in the opinion because this case --

No, no. But it's opaque for a very specific reason, because, you know, this case sort of piggybacks on the earlier BMG decision. And in the BMG decision, the Fourth Circuit says that once you have knowledge that the same customer or address is going to --

Well, I -- I think the customer did it as to the bill payer, the account holder --

-- which, of course, is the same unit that Cox uses when there's non-payment.

When -- when -- when Cox doesn't get payment from grandma, they don't --

I think that is the common law still. I don't think you need --

So, with -- with residential customers, I think it is, but in all events, that was ruled on in summary judgment in this case. And then what the Fourth Circuit did -- and I don't think you can read this part of the opinion differently -- they said that Cox forfeited any argument that they didn't have knowledge to the specific customers.

I don't -- what I would say is I think that's enough on the intent element to get to the jury. It's not enough to get directed verdict in my favor.

But it's enough to get to the jury. And Cox can go and they can try to argue that, oh, no, we had the purest intent, we tried really, really hard --

Yes. But, of course, I could point to the "F the DMCA" e-mail --

-- and I think I'd probably do all right on that exchange.

And, you know, keep in mind this went to a jury and it went to a jury of people who probably didn't want to lose their Internet connection. And yet, hearing all the evidence here, they had very little trouble saying this is a case where there is material contribution. This is a case where there should be aiding-and-abetting liability because of the conduct that took place here. And I guess I would just sort of take a step back from this to say, you know, if you think about this kind of in Halberstam or Twitter terms, you know, there's two kind of elements that are critical. One is the intent element. And if you look at Twitter and you look at Halberstam, the intent level is not purposeful. And then the other question, of course, is material contribution or, in aiding-and-abetting terms, substantial assistance. And I think the one way in which the copyright laws, the trademark laws, are a little bit different is I think they do think that when you are providing the means of infringement to somebody, I mean, that's pretty easy to show that that is material contribution, substantial assistance. And so, if you're looking to distinguish the FedEx's of the world, the electrical companies of the world, I'm going to agree with Mr. Stewart, I thought he did an excellent job of sort of parroting what my answer would have been with respect to the DMCA, but I would also say, you know, there's a difference when you're providing the means of infringement, which is what the ISPs do in these cases. That is going to be material contribution if you're doing it with knowledge of specific customers and what they're likely to do. And that's true in the trademark context as well, where you're providing it -- you know, the Inwood case is you had a drug manufacturer who's giving it to somebody and they know that person, that generic person is passing them off as branded pharmaceuticals, and that's where this Court said, yeah, there's liability for that.

I mean, there was, like, a separate question of proximate cause in that case that this Court didn't reach --

-- and I don't know the answer to, but I -- I think, you know, and -- and I think there were questions at oral argument about red-flag dealers and if you had specific knowledge you were providing them to the red-flag dealers. I think that probably would satisfy the -- the -- the -- the intent standard at least at the common law. Again, I don't know all of the details of the PCLAA or whatever it is that was at issue there, so I don't want to sort of speak against any of my other clients. (Laughter.)

But -- but I -- but I -- but I do think that with respect to the common law intent, that would be -- that would be satisfied.

So I -- I think they're wrong about that. I think they're wrong about that for reasons that are textual and also reasons that have to do with kind of the common law and the way that it treats aiding and abetting or secondary liability. So, if you think about the common law, you are -- if you have aiding and abetting, then you are essentially on the hook for the consequences of the mens rea of the person you're assisting and your own sort of mens rea beyond aiding and abetting doesn't count. So, if I aid and abet a first-degree murderer, I'm -- you know, either criminally or civilly, I'm on the hook for the first-degree murder and -- but whereas, if I aid and abet -- now all my conduct is exactly the same and I aid and abet somebody in manslaughter, the consequences are different. And then, if you look at the text here, the relevant text is 504 of the Copyright Act. It actually doesn't focus on the mental state of the infringer. It asks whether the infringement was willful. And, of course, in this context, it's the direct infringer who's doing the infringement.

So, I -- I mean, you know, my -- my -- my front-line answer would be I think the safe harbor takes care of the regionalized piece. And, frankly, I'm not that worried about the regionalized piece because, if -- if that were really the problem, we could go after the regionalized piece.

Right. And we could probably figure out who the regional ISP is and then go after its customers. That's why I'm not going to die on the hill of the regional ISPs. And I think you could -- I mean, you know, material support, I mean, you know, heaven knows this Court remembers from Twitter that, you know, by the time you look at the common law, you had six factors under Halberstam for what is substantial assistance. So there's clearly enough sort of -- if -- if you want to, if that's the -- if that's the hypo that's concerning you, then you can certainly sort of come up with a definition of "material contribution" that carves that out. I think the --

What's that?

And I'm inviting you to take my help.

I mean, I -- I don't want to, like -- you know, it's -- it's hard to just absolutely give it away in a case where that's not the way the case was litigated, but I am telling you that, you know, my -- my friend I thought was remarkably candid in admitting that if you adopt their position, that really is an extreme position. The DMCA and the safe harbors are a dead letter. If you carve out regional ISPs, the safe harbor is alive and well, and equally importantly, the incentives for both sides to come and to try to have a reasonable negotiation are alive and well.

Well, obviously, I prefer Door Number 3, but I would also prefer what --

Whatever the standard is, I'd prefer -- I -- I'd prefer a chance to pursue that. And I would think, even under purpose, there has to be room for treating somebody that says "F the DMCA" differently from somebody that tries their level best to comply and doesn't.

So let me take them in 101 turn. I mean, Henry says intent and purpose, but it also says you can infer that sort of compound if you know that the user is going to use it unlawfully. And, like, you know, I mean, to me, you know, the right way to look at this is consistent with the Re- --

It -- it -- it -- not maybe in the same line, but I think it says it elsewhere. At least that's what I take from the opinion. Maybe I'm -- maybe I'm misreading it.

But that's certainly -- that's certainly why we were relying on the -- sort of the Henry case, because you were providing it knowing it would be misused. So -- so -- so that's part of the answer on Henry. But, again, then you get to Kalem. And Kalem, yes, like, obviously, if you have inducement cases, that's what -- that's what advertising goes to. It goes to inducement 102 cases. But inducement cases are -- you know, historically, if you look at all of the contributory infringement cases, inducement is the smaller subset. And the dominant subset is material contribution, where you're providing material support, substantial assistance, went to somebody you know that it's an infringer. And that's why, like, you can get the people who are just packing and shipping the bootleg records, right? Like, you know, they don't, like, care what's in the box. They're just getting paid for doing it. But you still say and -- and always have said that they are liable. That's -- I mean, that's certainly the dominant common law rule. It's the rule in the trademark context as well. So let me just say a word about Kalem. Mean, Kalem, again, you know, is a case where you have, you know, the -- the promoter of the motion picture that's made from the book, so it's an easy case. And the language they talk about where Justice Holmes goes back to a couple of Supreme Judicial Court of Massachusetts opinions and talks about 103 indifference, the Court specifically says we don't have to get into those niceties here. So what they're relying on in Kalem, which is why we ignored it, is the purest form of dicta.

Thank you, Your Honor. Plaintiffs' test relies on specific known infringers who are certain to keep infringing. If Plaintiffs want to bring an action that is based upon specific known infringers who are -- who are certain to keep infringing, they can bring that case under their rule against ISPs. This Court cannot affirm on that basis, though, and neither can the courts below. Let's talk about how this case was 104 tried. There is no proof of any specific known infringer that is a human being who actually did any of the things that some of these hypotheticals spun out. Even if you focus on households, there is no way to disaggregate, precisely for the reason that Justice Gorsuch points out, Plaintiffs tried a case in gross as to 57,000 subscribers, no individual circumstances. Plaintiffs -- Plaintiffs never tried to disaggregate in any way as to the people who they see -- who they claim can sustain this verdict. Saying that the DMCA shows that there is liability in a world of uncertainty is not how this Court reads statutes. You can't infer liability from the blank page. And Congress said not to do that. The DMCA is also no panacea. ISPs have no certainty that they will keep the safe harbor. A lay jury decides whether an ISP has acted reasonably and whether terminations are appropriate. Justice Alito asked, what is an ISP supposed to do when confronted with a university? My friend gave an answer, have a 105 conversation. That's a terrible answer from the perspective of the company that is trying to figure out what its legal obligations are facing crushing liabilities. I have an answer. When Justice Barrett asked about recourse, the university -- the I -- the Plaintiffs have recourse. How about a conversation with the ISPs where they talk about how to work out things together? Maybe they kick in a little money. Now they won't get billion-dollar verdicts, but if they believe that the programs that Cox and others have aren't satisfactory, they can design better programs and help pay for them. If the Court has no further questions, we respectfully request that the court below be reversed.