Within the past 15 or so years, we’ve seen a large number of cases and laws addressing the issue of indirect liability — also referred to as secondary liability or third-party liability — for copyright infringement online.
In 1998, Congress passed the Digital Millennium Copyright Act, providing conditional safe harbors for online service providers. There’s also been a number of high-profile lawsuits against Napster, Aimster, Grokster, Limewire, and YouTube addressing indirect infringement.
Even so, there remains a misunderstanding among the public, especially online, about indirect liability. Why should X be liable when someone else is engaged in copyright infringement if X is only hosting/transmitting/providing a link to the unauthorized work?
This misunderstanding is embraced by TechDirt, who last week highlighted a story about an open letter from Russian ISPs to the entertainment industry in response to pressure on the ISPs for the copyright infringement of their users. TechDirt characterized the letter as explaining “basic liability concepts” to the entertainment industry. “The basic idea is that a third party service provider or tool should not be held liable for what users do with those tools,” explained Mike Masnick at TechDirt.
But that idea is not a “basic liability concept”; to the contrary, the idea ignores reality and the bulk of the history of law in most countries. Indeed, Masnick’s characterization of “basic liability concepts” isn’t even recognized in Russia1 — hence the reason the letter was written in the first place.
I previously wrote about secondary liability in Viacom Appeals Lawsuit Against YouTube, but I think the topic merits further discussion. There’s a gap in the common, public understanding of liability and how the law treats liability. The gap stems from a misunderstanding of tort law in general and is compounded when we start talking about indirect tort liability and how it is applied in the copyright context. You can see the gap in this understanding of tort law not only in debates concerning indirect copyright infringement, but also displayed in the perennial debates about tort reform, or in public reactions to what are sometimes referred to as “frivolous lawsuits.”2
To bridge the gap in understanding, I want to take a walk through tort law and why we have it. Next, I want to take a look at secondary liability in general and then specifically in terms of copyright. Finally, I want to look at how these concepts transferred to the online world and the legislative responses to dealing with the particular challenges of applying them to internet service providers. We’ll see that zero liability for service providers for the actions of their users is not a “basic liability concept” — far from being a logical or foregone conclusion, it is one of several possibilities of liability for service providers based on ancient concepts of tort liability, and it is an option that has, in part, been discarded in most of the world.
First, let’s talk about what “torts” are and why I brought them up.
These Torts are Delicious
Torts, in their broadest sense, are legal “wrongs”. They are private, rather than public, wrongs — distinguishing them from crimes — and they stem from universally applicable duties, rather than voluntarily agreed-upon duties — distinguishing them from contract claims.
The reason I’ll be talking about torts is that copyright infringement is a tort. In 1869, the Massachusetts Circuit Court wrote:
Rights secured by copyright are property within the meaning of the law of copyright, and whoever invades that property beyond the privilege conceded to subsequent authors commits a tort.3
This characterization of copyright infringement as tort has survived since then.4
The Goals of Tort Law
I think a big reason for the gap in understanding comes from a lack of understanding the goals of tort law. Most people understand the basic goals of criminal law. Some things are just bad — murder, robbery, rape. Society as a whole wants to either prevent them from happening or punish those who do them. Criminal law serves these goals; prison and fines act as punishment, the threat of prison and fines acts as a deterrent.5 The problem comes when people assume that the law of torts — private wrongs — has the same goals as the law of crimes — public wrongs. Deterrence and punishment are indeed part of the goals of tort law, but there are more goals to consider. One way of looking at it is laid out by Sir John William Salmond, who wrote in 1891:
A law is a rule of conduct set by the state and enforced by a sanction; a sanction is an evil inflicted upon those who disregard the rule; the administration of justice is that function of the state which consists in the infliction of such sanctions, that is to say, in punishment . Now, as a matter of fact, punishment is not an invariable or essential end of the administration of justice, but is merely one of several ends. These are at least three in number; namely, (i) prevention, (2) punishment, and (3) compensation.6
The addition of “compensation” to the ends of justice plays a large role in the administration of tort law. Indeed, nowadays economics plays just as much a role in shaping the law of torts as morality and ethics. Tort law professor F. Patrick Hubbard restates these goals from a modern perspective as such:
The tort system’s redistribution of the loss from the plaintiff to the defendant has been justified in terms of three policy goals. First, the liability for payment of compensatory damages prevents wrongdoing and thus protects rights in several ways, particularly: (1) the payment for injuries caused by wrongful conduct provides an incentive to avoid wrongful conduct; and (2) even where no wrongdoing is involved, imposing liability for accident costs provides an incentive to reduce injuries not currently preventable by due care by lowering the level of activity, or by seeking innovations that result in new, more cost-effective safety measures. Second, our sense of fairness requires that, as a matter of “corrective justice,” victims who suffer injury because their rights have been wrongly denied should have recourse to a system that requires injurers to pay compensation. These injurers “deserve” to bear the costs of their wrongs, not innocent victims. This concept of “just desert” also serves to limit liability from becoming disproportionately large in comparison to a defendant’s wrongdoing. Third, compensation of victims is frequently said to be, by itself, a goal of tort law.7
A thorough discussion of the goals and justifications of tort law is well beyond the scope of this article; for more, check out Stanford Encylopedia of Philosophy’s entry on Theories of Tort Law. For purposes of our discussion, it’s enough to say that modern tort law goes beyond merely “placing blame.” It asks also, among other things, what are the most efficient ways to allocate the losses attached to wrongs among the parties involved and what rules provide the best incentives to reduce the likelihood of wrongs. While differing theories about tort law compete, it is safe to say that the goals of “fairness” and “efficiency” are at the heart of modern tort jurisprudence.
Secondary Liability in Tort Law
The gap in the public’s understanding of tort law grows wider when it comes to secondary liability — those doctrines that place liability on someone other than the person who actually committed the wrong. How can this be? In 1883, political economist William Edward Hearn wrote:
§ 4. The general principle of justice which governs men’s responsibility for their conduct has two branches. One is that, subject to certain specified grounds of defence, every man is answerable for his own acts and his own forbearances, or for those which he has ordered. The other is that no man is answerable for the conduct of any other person. To the latter proposition, however, there are certain exceptions. These are cases of vicarious liability.8
Vicarious liability grew out of the long-established doctrine of respondeat superior in common law. Respondeat superior states that an employer, or principal, is liable for the acts of his employee, or agent.9
Contributory liability is the name most often used in the intellectual property context to describe the doctrine of holding someone liable for their contribution to the commission of a tort.10 The theory behind it is sometimes referred to as “concert of action” and it finds its genesis in the criminal law idea of “aiding and abetting.”11
“The most ancient authorities of the law” considered those who assisted in the commission of a crime to be guilty of the crime itself.12 So, for example, a getaway driver for a bank robber could be found guilty of the robbery even if his role was limited to sitting in the car while others did the actual robbing.
This principle remains the same when the wrong in question is a tort rather than a crime. In 1831, the Supreme Court of Ohio wrote in Bell v. Miller:
All concerned in the commission of a trespass are considered principals. An assault and battery may be committed by a party not present, if he be a principal actor in or adviser and promoter of making the attack. If one person employ another to commit an assault and battery or any other trespass, and the act is perpetrated, both are guilty, and both responsible in damages. It was not supposed that this was now a debatable question.13
The justifications for secondary liability build off of the fairness and efficiency goals of tort law. Associate professor of law Mark Bartholomew restates it as such: “While vicarious liability exists to redistribute risk in the absence of fault, contributory liability seeks to apportion liability on the basis of moral dessert.”14
Secondary Liability in Copyright Law
As mentioned above, copyright infringement is a tort. One would expect, then, that doctrines of secondary liability have made their way into cases dealing with copyright infringement. And indeed, they have.
Courts have generally recognized three theories for holding a third party indirectly liable for direct infringement caused by another:
- Vicarious infringement — where one profits from direct infringement that one has the right and ability to control15
- Contributory infringement — where one has knowledge of the infringement and causes or materially contributes to the infringement16
- Inducement — where one engages in purposeful conduct that encourages infringement with the intent to encourage infringement17
These well-established principles of secondary liability have been applied in infringement cases for decades, if not longer.18 But it wasn’t until the widespread adoption of the internet over the last 20 or so years that these doctrines have played an increasingly important role in copyright infringement cases.
One quick note: some copyright critics cling to the belief that secondary liability doesn’t exist in the realm of copyright infringement because it is not mentioned in the Copyright Act. The belief is bolstered, perhaps, by Justice Stevens’s remark in Sony v. Universal City Studios that “The Copyright Act does not expressly render anyone liable for infringement committed by another.”19 But the idea that third-party liability in copyright infringement is invalid is wrong on two counts.
First, the US operates under a common law system. The law is developed in the courts unless and until it is modified by statute. As shown above, courts have long recognized doctrines of secondary liability in infringement cases. And if you read just a couple sentences more, you see Stevens explain the same:
The absence of such express language in the copyright statute does not preclude the imposition of liability for copyright infringements on certain parties who have not themselves engaged in the infringing activity. For vicarious liability is imposed in virtually all areas of the law, and the concept of contributory infringement is merely a species of the broader problem of identifying the circumstances in which it is just to hold one individual accountable for the actions of another.
Second, even recognizing that secondary liability exists in copyright law under the common law, some have suggested that Stevens was, in fact, wrong that the Copyright Act didn’t expressly recognize it. Peter Menell and David Nimmer explain:
The extensive legislative history underlying the 1976 Act refers specifically to the law of torts in clarifying that “where the work was infringed by two or more joint tort feasors, the bill would make them jointly and severally liable …” More references to tort law undergird the 1976 Act’s remedial provisions. These references cement the proposition that tort doctrine furnishes the background law for determining what circumstances render someone liable for infringement and, if liable, the scope of remedies.
The legislative history makes two direct references to indirect liability standards, both of which support the continuation of then-existing doctrines and their further refinement through judicial decisions. In explaining the general scope of copyright, the House Report recognizes contributory liability:
The exclusive rights accorded to a copyright owner under section 106 are ‘to do and to authorize’ any of the activities specified in the five numbered clauses. Use of the phrase ‘to authorize’ is intended to avoid any questions as to the liability of contributory infringers. For example, a person who lawfully acquires an authorized copy of a motion picture would be an infringer if he or she engages in the business of renting it to others for purposes of unauthorized public performance.
In discussing infringement, the House Report explains:
Vicarious Liability for Infringing Performances
The committee has considered and rejected an amendment to this section intended to exempt the proprietors of an establishment, such as a ballroom or night club, from liability for copyright infringement committed by an independent contractor, such as an orchestra laeder [sic]. A well-established principle of copyright law is that a person who violates any of the exclusive rights of the copyright owner is an infringer, including persons who can be considered related or vicarious infringers. To be held a related or vicarious infringer in the case of performing rights, a defendant must either actively operate or supervise the operation of the place wherein the performances occur, or control the content of the infringing program, and expect commercial gain from the operation and either direct or indirect benefit from the infringing performance. The committee has decided that no justification exists for changing existing law, and causing a significant erosion of the public performance right.
That excerpt shows legislative intent to preserve the principles of vicarious liability that had developed through the courts under prior law under the 1976 Act.20
Copyright Liability Online
With the advent of digital technology — allowing text, music, and visual works to be reproduced easily and perfectly — and the internet — allowing digital files to be distributed cheaply across the globe — liability for copyright infringement quickly entered new territory.
Copyright infringement has traditionally been characterized as a strict liability tort, meaning one can be liable without intending to infringe any of the exclusive rights protected by copyright.21 This strict liability, in theory, implicates many parties that make up the backbone of the internet since transmission of digital data necessarily involves “copying” and distribution, as well as public display or performance, at many different stages.
In the US, the Digital Millennium Copyright Act (DMCA) — specifically, the safe harbors the bill amended to §512 of Title 17 of the US Code — was a watershed bill for dealing with copyright liability online. One can divide how the law treats copyright liability for online service providers into two eras: pre-DMCA and post-DMCA.
Prior to the DMCA, it was up to courts to determine whether service providers were directly infringing on copyright for their role in copying and transmitting infringing works originally uploaded by users of the service. In at least one case, Playboy Enterprises v. Frena, the court found that the service provider was liable for direct infringement.22 Despite Playboy Enterprises, however, the case law suggests that courts were largely unwilling to hold service providers liable for direct infringement.23
When we turn to the question of indirect infringement — whether a service provider is liable for vicarious or contributory infringement — the pre-DMCA case law shows a far more muddied application of the doctrines. Contrasting approaches to these doctrines as applied to service providers emerged in the courts, with the result being a great deal of uncertainty for service providers.24
Governments around the world faced the same issues. The internet represented a great source of innovation and advancement. How best to craft liability rules that would ensure continued innovation and advancement while protecting long-established, real world rights online? The US Congress ultimately crafted the DMCA to address these concerns and provide certainty.
Addressing Online Liability Through Legislation
The choice for what level of liability online service providers should be subjected to, by law, generally comes down to one of three options: no liability, strict liability, or some type of limited or conditional liability. The DMCA takes a mixed approach, depending on what role an online service provider has taken on. Service providers have zero copyright liability if they are only engaged in transmitting and routing data, or if copies are made for system caching purposes.25 Conditional liability is placed on service providers for infringing works residing on the provider’s system at the direction of a user — for example, an unauthorized video uploaded to YouTube, an unauthorized photo posted on a message board, or an unauthorized song hosted on a user’s web site.
The general consensus is that conditional liability for service providers in the copyright context is the best of the three options.26 This approach balances the multitude of competing factors involved: it creates incentives on parties to reduce the harms of copyright infringement while avoiding overzealous monitoring that may suppress noninfringing speech. Most major countries around the world have adopted some form of conditional liability for online service providers in the copyright realm.27
What would liability for service providers look like today had there not been a DMCA? The answer is unclear. Certainly, there would have been a period of great uncertainty at the early stages of the growth of the internet as courts addressed the question. Perhaps the common law would have developed toward standards which fairly and efficiently allocated the risks of user infringement among the various parties; it is not unlikely, however, that courts could have adopted rules placing more of a burden on service providers for user infringement. Either way, if the question was left to the courts, it would likely have stifled innovation when such a result would have been most detrimental. A large company at the time like Yahoo or AOL could easily fend off legal challenges, where a cash-strapped startup might have folded, unable to fight in court no matter how likely its eventual success would be.
Zero liability for online service providers is far from a “basic liability concept.” Tort law has long recognized several doctrines where indirect liability may be imposed, and copyright law has long adopted those doctrines. Absent legislation like the DMCA, it is uncertain how courts would have applied those doctrines to service providers in the online context. The period of greatest uncertainty in the law would have coincided with the early stages of the widespread adoption of the internet: a time when uncertainty would have provided the most detriment.
It’s true that the laws regarding online liability for service providers are far from perfect. Stakeholders on all sides have expressed areas where they could be improved. But claims that zero liability for service providers are a foregone conclusion from a “common sense” point of view or that they reflect “basic liability concepts” are entirely off-base.
- Douglas Lichtman and William Landes, Indirect Liability for Copyright Infringement: An Economic Perspective, 16 Harvard Journal of Law & Technology 395 (2003) — an article that provided much help to me in shaping this article; it is well worth a read for a broader look at the topics addressed here.
- Christopher Osakwe, Russian Civil Code. Parts 1-3: Text and Analysis (Wolters Kluver 2008). [↩]
- Most notably the “hot coffee” lawsuit against McDonalds. [↩]
- Lawrence v. Dana, 15 F. Cas. 26, 61 (C.C. Mass. 1869). [↩]
- Peter S. Menell & David Nimmer, Unwinding Sony, 95 California Law Review 941, 996 (2007). [↩]
- This is obviously an over-simplification for the purpose of discussion. For a more detailed look at theories of criminal law, start at Stanford Encyclopedia of Philosophy – Theories of Criminal Law. [↩]
- Essays in Jurisprudence and Legal History, pg. 123. [↩]
- Hubbard, The Nature and Impact of the ‘Tort Reform’ Movement, 35 Hofstra Law Review 437, 445-46 (2006). [↩]
- The Theory of Legal Duties and Rights, pg 125. [↩]
- See, for example: Broom, Manisty, and Cagney, A Selection of Legal Maxims, Classified and Illustrated, pg. 798 (1884) for a discussion of the doctrine and its long-standing place in history. [↩]
- For example, in trademark cases like Inwood Labs v. Ives Labs, 456 US 844 (1982) and especially in patent cases: Thomson-Houston Electric v. Kelsey Electric, 72 F1d 1016 (1896). [↩]
- See In re Aimster Copyright Litigation, 334 F.3d 643, 651 (7th Cir. 2003); Boedecker, Kasulis, Morgan, and Stoltman, The History of Enterprise Liability, Ninth C.H.A.R.M. Conference on Historical Analysis and Research in Marketing, 205, 210 (1999). [↩]
- The American and English Encyclopedia of Law, pp. 29-30 (Edward Thompson Company, 1896). [↩]
- Bell v. Miller, 5 Ohio 250 (Supreme Court of Ohio, 1831). [↩]
- Mark Bartholomew, Copyright, Trademark and Secondary Liability after Grokster, 32 Columbia Journal of Law & the Arts 445, 465 (2009). [↩]
- Shapiro, Bernstein & Co. v. HL Green Company, 316 F.2d 304, 308-309 (2nd Cir. 1963). [↩]
- Gershwin Publishing Corp. v. Columbia Artists Man., 443 F.2d 1159, 1162 (2nd Cir. 1971). [↩]
- Arista Records v. Lime Group (SDNY 2010). While there has been some doubt as to whether inducement is a new theory of secondary liability or just a restatement of contributory infringement, the court in Arista Records concluded that the Supreme Court in MGM v. Grokster confirmed inducement as a “distinct cause of action.” [↩]
- See Unwinding Sony, pp 996-1005 for a thorough discussion of the history here. [↩]
- 464 US 435. [↩]
- Unwinding Sony, pp. 995-96. [↩]
- Patry on Copyright, §9:5. [↩]
- 839 F. Supp. 1552 (M.D. Fla. 1993). [↩]
- Alfred Yen, Internet Service Provider Liability for Subscriber Copyright Infringement, Enterprise Liability, and the First Amendment, 88 Georgetown Law Journal 1833, 1842 (2000). [↩]
- Yen, 1877. [↩]
- 17 USC § 512 (a) and (b). [↩]
- See, for example, Alan Woodfield, When Should the Bell Toll? The Economics of New Zealand’s Debate on Indirect Liability for Online Copyright Infringement, 1 Review of Economic Research on Copyright Issues 119, 147 (2004), discussing why New Zealand’s adoption of conditional liability was the best of the three options. [↩]
- See Woodfield; Aditya Gupta, The Scope of Online Service Providers’ for Copyright Infringing Third Party Content Under the Indian Laws – the Road Ahead, 15 Journal of Intellectual Property Rights 35 (2010) (discussing UK, Canada, Australia, and India); VK Unni, Internet Service Provider’s Liability for Copyright Infringement – How to Clear the Misty Indian Perspective, 8 Richmond Journal of Law & Technology 13 (2001) (also discussing Canada, Australia, and India, as well as Singapore); the procedure the Russian ISPs propose in their letter highlighted in the TechDirt article mirrors these approaches. [↩]