IPSC Breakout Session 1 Copyright Secondary Liability

Copyright’s Intent, Mark McKenna (with Laura A. Heymann & Alfred C. Yen)

Cox: Contributory infringement for service providers can be shown only with intent, or with a service tailored to infringement. What happened to Gershwin? Will courts really accept that? Will they expand vicarious liability more to capture intuitions they have about fault?

Common-law tort origin is the claim, but ©’s rules are misaligned from tort. Is this all © exceptionalism? Many of us feel that courts are searching to attribute fault but don’t have the right anchoring in tort law fault principles.

First, ©’s insistence on hard line b/t direct infringement and secondary liability. Tort law more flexibly assigns first-party liability to those whose conduct causes the injury—co-tortfeasors when contributions are deemed sufficient, even when another party’s participation is required for harm and even when co-tortfeasor’s conduct is not the proximate cause. Joint liability—tort doesn’t regard that liability as secondary. They’re not derivatively liable for wrongful conduct of another, but liable b/c of their own wrongful contributions to harm. If party most proximate to harm is batterer, the co-tortfeasor is also a batterer. Tort only very rarely imposes truly secondary liability (wrongs committed by others). Vicarious liability is the standard, but it’s not specific to the particular conduct but rather exists because of the broader relationship (employment) to the tortfeasor. Enterprise liability too.

Basically true even for strict liability like blasting/harboring a wild animal. Even when D’s own conduct is also sufficient to subject them to liability—when landlord engages w/wild animal in ways that could be called a harborer—that’s not derivative liability. Otherwise landlord might be negligent in their own conduct (renting to a known harborer). So © being strict liability doesn’t demand a different approach.

Framing of whole Q of secondary liability is thus already off on the wrong foot from tort perspective.

© ignores wrongful contribution to an injury: party’s own negligent conduct exposes the P to third-party wrongdoing, like landlord who exposes tenants to assault, or train leaving passenger in dangerous area. Liability there is not co-tortfeasor; this category doesn’t put D in same category as other tortfeasor: negligence liability, not battery, for negligently exposing P to third-party battery. Doctrines like causation, act/omission, etc. come into it. The remedies are negligence remedies—no punitive damages even if third party engages in intentional tort.

How does © get misaligned? Courts initially concerned with co-infringers—people who performed part of the act that led to infringement. Drawing on patent cases where infringement requires assembly of lots of parts. A common enterprise, either directly or indirectly where one party makes a component to be combined with another. When co-infringers weren’t amenable to suit, Ps tried to bring in more upstream participants to say they were also co-infringers. Courts started using the term “contributory infringement.” But they weren’t really developing it as a theory of derivative liability, but rather describing why the contributions of Ds made them co-infringers. These were common enterprises—common purpose to cause the infringement.

Part of the confusion in © is emphasis on secondariness of liability, unplugged from fault principles. If we tried to do more faithful mapping to tort law: we should be very reluctant to impose true secondary liability other than vicarious liability, which we would define much more narrowly—a relationship like employer/employee or joint enterprise—relationships not specific to the infringing conduct, with much higher levels of control. Thus, we’d reorient to (1) co-infringers—liable b/c their own conduct makes them fairly labeled an infringer, subject to © remedies, including inducement and providing products w/no substantial noninfringing use; could also include, per Sotomayor, other cases of aiding and abetting where there’s knowledge plus material contribution as in tort law where tort law demands significant contribution at a time when it can be said to reflect a meeting of the minds in furtherance of tortious conduct, not just any time/knowledge of past infringement. (2) negligent exposure to third-party infringement; implication of duty, breach, causation and damages. Not really © infringement but negligence, so the damages are different.

Lea Bishop: so is © not really a strict liability offense?

A: these doctrines aren’t limited to intentional torts, but the way the courts talk about the other party’s contribution is about intent to interact w/other D—so it’s not intent in the “intended to infringe” way but “intended to work with.” Underlying tort doesn’t require proof of intent. Cox’s categories of intent map pretty well if they mean “intent to work together” rather than “intent to carry out infringement.”

The New Law of Vicarious Liability in Copyright, Michael Carroll

Cox tells courts to change their vocabulary. Does that matter? If so how? Restatement (Second) of Agency conceives of two types of principal/agent employment relationships. Master/servant where there’s agency v. independent contractor who may or may not be an agent. It was against that tort law background that the 2d Circuit decided Shapiro, Bernstein about whether a department store owner/operator was liable for sales of infringing recordings by concessionaire. Court rejects independent contractor defense: right and ability to supervise plus obvious and direct financial interest in exploitation of copyrighted materials—looking to © policy and says policy is best served by imposing liability in these circumstances. Courts repeat those terms, not with full consistency.

Sony came along: The use of the term vicarious liability was imprecise in the opinion, but Justice Stevens used the term for any kind of indirect liability. 9th Circuit kept treating Grokster as simply ratifying Shapiro, Bernstein. But Grokster’s formulation is different! Ignored Justice Souter’s restatement; Cox says that the Court is the boss. You’ve got to start using the magic words. If one infringes by profiting from direct infringement, financial interest isn’t enough: profit is revenue minus cost. But maybe it’s not a big deal. Declining to exercise a right to stop or limit direct infringement: to decline to take action implies both knowledge and intent. But as an outgrowth of respondeat superior, vicarious liability has been treated as strict. How can that be reconciled?

Cox’s treatment of Grokster suggests that the Court will rely on its own restatement of vicarious liability when it reaches a relevant case. Courts will need to develop a standard for declining to exercise right/ability. A relationship akin to employment will impute knowledge to hold that supervising party declined to act; but a service provider for an internet user would require more evidence than a contract that said you could be cut off w/o some specific notice.

Example: tape machine manufacturer & its retailers: 8th circuit said they were liable b/c of contracts w/retailers for how machines were used; machines were distributed free; some tape sales were to infringers & manufacturer profited from tape sales. Not sure that could suffice.

Question: where did Souter get his words? Not in the briefs in Grokster.

Eric Goldman: thought that Grokster was p-favorable; what’s the empirics?

RT: wouldn’t it be stronger to start from the point that the real name of/justification for vicarious liability is agency liability? Service user is obviously not agent of service provider, so that would help realign with larger agency/vicarious liability law.

Also: read the SCOTUS sexual harassment cases from 1998 & Ginsburg’s characterization of the liability standard, where she makes similar moves about vicarious liability.

Grimmelmann: as with shadow docket, SCOTUS is telling lower courts to read their tea leaves and comply.

Copyright Exceptionalism in the Supreme Court’s Secondary Liability Cases, Pamela Samuelson

© industries seek broad liability rules for tech companies; industry turned to Gershwin definition from 2d Circuit in Sony, Grokster, & Cox, claiming that Gershwin was the bedrock foundation for their claims. © exceptionalist arguments derived neither from statute nor common law, but claims about massive uncontrolled infringement. But Ct even in Sony didn’t cite to Gershwin except as vicarious case.

Grokster: MGM no longer relying on Gershwin alone, but emphasized that Gershwin’s definition of contributory liability included inducement. Court looked to Gershwin for inducement as well as to patent law. Cox is a lazy opinion; Sotomayor is making more sense. Asks: why not aiding & abetting? Some options: there’s nothing in the statute; the statute says “to authorize” which didn’t happen; no inducement b/c Cox didn’t encourage infringement; no special tailoring, no direct financial benefit; no a&abetting b/c there’s no intent to aid infringers. Material contribution w/knowledge was Sony’s only chance, but broadband service wasn’t proximate cause and Cox’s after the fact knowledge/lack of way to know which user actually infringed was too limited to justify liability.

Pressure on vicarious infringement & volitional conduct will exist, but probably not on inducement b/c Hikma reinforced the requirement of active inducement in patent context.

Did SCt really intend to overturn Netcom such that failure to take something down after notice is no longer material contribution sufficient to justify liability? DMCA is not a dead letter b/c many incentives to comply still exist. [Including incentives to comply for other countries’ regimes.]

Jim Gibson: Even Sotomayor wants to use a&a for a heightened mens rea—intent of helping other person succeed in committing wrongful conduct, not just intent to perform the act that enables infringement. So the common focus on intent seems like the most limiting factor going forward, not whether a&a can also be included.

A: Taamneh was important to Cox despite few mentions—the required intent (to aid terrorists) was something the Court thought about in that context. She expects intent to get watered down.

Fearing (and Loathing) the Common Law of Copyright, Shyam Balganesh

Why the reluctance of the Court to engage with © as a common law system? Thomas says: we’ve recognized specific forms of secondary © liability that predate the Act, but we’re loath to expand liability beyond that. Sotomayor says: why?

Legislative-judicial dynamic around parts of the statute has been essential to the 1976 Act. Typology: Legislative modality: novation; judicial task: interpretation (Congress invented this and didn’t draw from prior case law); example: joint works

L: Codification; J: preservation/interpretation; E: first sale

L: Silence (decided not to speak; mess already existed by 1960s & 1970s); J: unconstrained law-making; E: substantial similarity

L: Delegation; J: constrained law-making; E: fair use.

Sony has the right result but made a methodological error: Stevens says that Act doesn’t expressly render anyone liable for infringement committed by another & talks about absence of express language requiring courts to jump in. This is only partially true b/c of “to authorize” in 106, which Sony & Cox don’t mention—it has clear instantiated meaning and long history. Thomas has a theory of congressional primacy; the legitimacy of court-made law is always tested against backdrop of congressional action or inaction; seemingly a majority of the Court has accepted/acquiesced to this view. Clear patterns in his opinions in Star Athletica, dissents in Public.resource.org and Oracle: you’re using fair use to annul the statutory treatment of software. Even in Fogerty v. Fantasy: text of statute is clear; interpretation ends.

Thomas’s disdain for common law; three views. If there’s a backdrop of rules against which Congress legislates, appealing to those rules is legit but frozen in time. For delegated lawmaking—ongoing elaboration required for open-ended terms with express or implicit recognition that judge-made law will follow—he thinks it’s legitimate if the text constrains it with guardrails; securities law is an example of his objection b/c there’s not enough guardrails. Finally, independent lawmaking is wholly illegitimate (no gov’t edicts doctrine).

Maybe this was a category error: failing to discuss “to authorize.” Doesn’t think so, though, b/c briefs raised it. But he wasn’t convinced that there was a textualist hook. Raises Q: what does this do to other parts of © law if this vision of interpretive structure has a majority? Fair use implications: only if it is compatible w/the rest of the statute. Originality: same plane. Infringement analysis: implications for legislative reform. Beware of textualists when advocating for reform: how a court would handle that.

RT: Textualism masks that placement in categories is contestable: Glynn Lunney: reproduction/derivative works could have replaced substantial similarity; codification could have been read as novation in interpreting first sale versus exhaustion.

NO FAKES and similar ROP proposals often have language like “to the extent protected by the 1A” in their exclusions—how could this form of textualism handle that? Would Thomas’s approach ignore those exclusions just like he ignored the open ended language about useful articles in Star Athletic.

A: irony of Thomas’s MO: claiming that there’s plain meaning while refusing to look at legislative intent—he ignores “to authorize” b/c explaining what its plain meaning was would require a citation to the legislative history.

Our Byzantine Secondary Infringement System, James Grimmelmann

Conventional view: in US, there’s vicarious infringement and then intent-based contributory liability. He wants to do a thorough survey of all the secondary liability doctrines in US law, describe & critique it as a system, and then possibly suggest fixes.

True secondary liability doctrines: liability for someone else’s completed act of direct infringement.

What about infringement by authorization? Issuing a purported license w/o the right to do so—seems literalist but probably killed by Subafims.

Agency law: respondeat superior is used all the time where companies are held liable for employee’s actions. Agent’s actions and knowledge are imputed to their principal, often invisibly, even when not actuated by purpose to serve employer [not sure this last is true—looking forward to cites]. Especially in PRO licensing cases where employees at a bar are used to hold owners liable even when corporate law wouldn’t do it.

Volitional conduct/the server test: these often cut in opposite directions and interact weirdly with licenses granted to platforms by users.

Quasi-secondary liability: for conduct that could facilitate infringement regardless of whether there is actually infringement.

Scaffolding doctrines: direct infringement has no mental state requirement; makes stakes much higher for direct/secondary. Willful & innocent infringement also matter to statutory damages.

Criminal liability for willful infringement; brings in general criminal doctrines of aiding & abetting; there’s also a “causing” criminality but no federal attempts criminal liability. There’s also conspiracy liability: it’s a crime to conspire to criminally infringe; Pinkerton: conspiracy to commit any crime subjects conspirator to liability for any criminal infringement that’s reasonably foreseeable and in furtherance of criminal conspiracy. RICO: © infringement is a predicate crime.

512: Does it displace common law? Volitional conduct? Apply beyond enumerated services? Courts generally say no to all. Tony Reese has given good reasons to think it’s a bit more complicated. The exceptions it carves out all sound in secondary liability (quasi-contributory; quasi-vicarious—presumed that direct liability wasn’t possible so how could it be the same as common law liability?); what about the repeat infringer suspension—what is a reasonable policy? Recreated a lot of secondary infringement doctrine under the head of 512.

TPMs also matter: Serial copy management systems—you must implement them and you’re liable for distributing tech w/o them—that’s a kind  of quasi secondary liability.

1201 is too, arguably mapping onto Cox intent prongs—distributing tech “primarily designd for,” knowingly marketed for use in circumvention, or have limited commercial use except for circumvention.

1202 is too: knowingly language but not in any coherent/organized fashion.

This is far too complicated. There are way too many minor variations and overlaps. What happens to 512’s quasi contributory liability exception now that Cox has repudiated knowledge plus material contribution? Overlapping but inconsistent tests. Confused relationship of statutory codification to common-law elaboration: volitional conduct, server test, and 512 all seem to do similar work. Loopholes and traps for the unwary—Aereo was $100 million waste.

Jim Gibson: distribution liability can be thought of as secondary liability for the underlying reproduction, though the statute doesn’t say that.