Session 3: Copyright Doctrine
Patrick Goold , UC Berkeley School of Law (fellow)
Is Liability for Copyright Infringement Strict?
Strict liability is conduct plus outcome. Fault-based is conduct, outcome, and fault. Fault can be based on the standard of conduct (reasonableness), or the state of mind (intentionality, recklessness). Fault in the action or the actor.
With copyright, the conduct is copying, the outcome is substantially similar work. Doesn’t require culpable mental awareness. But is the result conditioned on the D failing to comply with a standard of conduct? His answer: yes, that’s what fair use is for.
Under usual fault based torts, burden is on P to prove existence of each element, but in © only conduct and outcome must be proved and burden is on D to prove absence of fault. That’s why he uses res ipsa loquitor: there are situations where the law presumes fault, and that’s what’s happening in ©.
Ochoa: Fault proves very sticky. Products liability—courts use the concept of defect to reintroduce fault into the equation. Can sue everyone in the chain of distribution, even people who weren’t responsible for the defect. In ©, same thing: can sue everyone who reproduces/distributes an ad in a print magazine even if they weren’t responsible.
A: Relationship-based strict liability also exists, vicarious liability. Similar issue.
In tort, whether conduct has an outcome has many variables—accidents can be caused by external factors. In copyright, relation between conduct and outcome seem to be entirely based on conduct. He’s thinking about ideas of fairness/reasonableness in negligence & their relation to this idea of strict determination. (Seems to me a question of how you define “conduct,” at least in substantial similarity cases—if “conduct” at issue is “writing a play,” then many variables do seem to matter.)
Brad Greenberg, Kernochan Center for Law, Media and the Arts, Columbia Law School (fellow)
The Uneasy Case for Teachnology Neutrality: Copyright Future-Proofing and Other Myths
Idea of media neutrality—“fixed in any tangible medium of expression.” Idea of future-proofing the statute. (Hi, Aereo!) Also to avoid disputes among incumbents and users—know what you have to pay for. Also to avoid need to redo the law, as Congress had to do a bunch before 1976 Act. DMCA was the major change after that, with smaller amendments; now trying again.
Paper is about why this failed. Conferred broad rights, and courts responded with lots of carveouts, which undermined the basic goals of tech neutrality. Applying rules/standards literature helps illuminate this. Tech neutrality isn’t a standard. It’s just a very inclusive rule, with anything new getting sucked in—overinclusiveness is a hallmark of a rule, not a standard. Courts have responded with convergence, moving rules towards standards, losing certainty.
Problems: disrupts incentive system if you believe ex ante incentives are important. Betamax, Perfect 10, Cablevision. Promotes inefficient risktaking for technologists, leaves copyright owners griping about stuff that was supposed to be covered.
Implications: if we want to future proof the statute, we actually need narrow rules to start and easier process to add more rules for certainty.
RT: incentives point only true if you believe incentive responses are finely calibrated. I don’t think the empirics on that are solid. Is Europe different/better?
A: hasn’t figured out best comparison—Canada is moving to tech neutrality. Perfect 10 is a perfect example: to keep Google from being liable, court crafted very broad transformativeness test. Tech specificity could have done much better. But hasn’t figured out best way to handle this case if there were no rules to start with. (Hmm. Much better than what? What have the negative consequences been? Because I bet many of us would consider them positive consequences. That’s why I asked about Europe, which is now trying to figure out if it needs fair use to give flexibility. We sure seem to be outperforming other systems that are also fighting about search engines.)
Lunney: Aereo? Tech neutrality is fine, but it’s equivalent to 1000 antennae on 1000 homes, but also equivalent to a cable system—which is it? Economically, if anyone could set up to compete with them, they should be more like antenna installers; if they’re a monopoly, more like cable.
A: Needs to work through that. © statute doesn’t handle private use very well generally.
Q: what are you envisioning © would do instead of being tech neutral? Congress takes a while to act.
A: This is where an admin agency could help. (RT: Joe Liu has an excellent article on this.)
Q: you say you want narrower rules and more rapid decisionmaking. How important is narrowness if you have administrators act quickly? If presumptively everything is protected, but then an agency can allow it.
A: problem with broad rule is asymmetry in adversarial interests. Challengers won’t have as much of an interest as incumbents. Public choice!
Garcia: history of rules that are too specific—webcasting. It’s not just the tech, the business model. Ad-supported streaming services: do they fit under the narrow statutory license? Business model/consumer preference neutrality is the killer issue.
A: maybe he wants industry-specific rules.
Ochoa: admin agencies are notoriously subject to capture. Why any better than Congress?
A: Some of this is unavoidable.
Buccafusco: one-way ratchet is the problem.
A: that’s what fair use has been forced to do. Goal is to take some of the stress off of fair use.
RT: this is a point of pushback: some of us think fair use is not overstressed, and this fits into a bigger political conversation about whether other nations would benefit from fair use.
A: not interested in that here, but understands its place.
Said: might help to decouple normative and descriptive. Feels as though you’re advocating before you worked out implications. Free yourself to figure out your descriptive account, and explore tech neutrality pros and cons.
D.R. Jones, The University of Memphis Cecil C. Humphreys School of Law
Law Firm Copying and Transformative Fair Use: An Examination of Different Purpose
Examines a group of cases with unlikely defendants—law firms. Publishers sued for infringement. Two wins: both hinged on transformative fair use, focusing on defendants’ purpose. There were no changes/alterations in works. A number of similar cases: plagiarism software, Perfect 10, Google Books—purpose can be transformative. These patent cases fit in.
Initially transformativeness wasn’t very clear, but a number of studies have identified patterns in what courts consider. Purpose, regardless of change in substance of work, is dominant.
There still seems to be some confusion—is purpose part of transformativeness, or is it something separate? “Creative metamorphosis.”
Goals of copyright: furthering creativity, research, expansion of knowledge—see this in search engine cases, plagiarism, and also the patent cases. There’s almost a presumption of fair use when you use a copyrighted work as evidence. Commentators, legislative history say this; very little analysis. Why would that be so obvious? Unless work is created to be used as evidence, it’s fair use. Two “evidence” cases talk in more detail: use as evidence is different purpose. Courts also say there’s a social interest in having all relevant information presented. Copyright must yield to the need for a complete evidentiary record; © can’t be used to undermine truth-seeking function. Something more important than ©. Societal benefit as a fundamental reason for fair use: supporting other values. You don’t necessarily need to create something new, as the evidentiary uses don’t. Societal benefit can trump need to pay. Even copies made for files/clients and not sent to the patent office were part of this important system, and thus also transformative. May not need to use transformativeness.
Lunney: maybe we’re using the term transformativeness too broadly. Is that a problem? Are cases coming out wrong?
A: maybe it’s semantic frustration. Doesn’t think an outcome change is needed. But forces courts to say weird things when the underlying reasoning is solid.
RT: Societal benefit doesn’t answer the Q of whether the P should get paid. My speculation: because factor four had become circular, courts turned to factor one to protect people creating new value when Ps were only making the Texaco argument that they’d like to be paid. In such cases, courts found transformative purpose dispositive. This move was especially likely when the new uses depended on multiple inputs and a potentially significant orphan works or holdout problem (true in Perfect 10, also true in the patent cases). Judicial proceeding/evidentiary use cases may be different from the patent cases in that the evidentiary uses are classic Wendy Gordon fair uses; the P is usually trying to protect something other than an economic interest.
A: it is clear there is no market or no effect on the market in evidentiary use cases. It’s really the fourth factor under all that. But in patent cases the courts turn to public benefit as a reason to reject Texaco. There has to be a social benefit that can outweigh even the Texaco argument that D can afford to pay licensing fees.
Loren: Semantics matter here. Several different practitioners have lamented how awful transformativeness is in explaining fair use to clients—same term, lots of different uses. To have new terms to support existing results could be useful in advising or explaining. (RT: Or Fair Use Best Practices!)
Ochoa: problem is the history. Productive use is much closer to what we’re talking about, but SCt shot that down, so it’s not surprising that courts won’t use that label.
Q: even you as member of the public might be confused and possibly deterred from making certain uses out of fear. Once you extend this reasoning and try to use other terms, you could be addressing some sort of broader societal chill. (RT: Or we could try Fair Use Best Practices! OK, I admit, I have a strong bias here.)
Peter Yu, Drake University Law School
Can the Canadian UGC Exception Be Transplanted Abroad?
Hong Kong consultation on copyright reform. Political parodies/using © work is common, and you can’t realistically call Beyonce for permission.
Option 1: clarifying criminal provisions; Option 2: specific criminal exemption for parody; Option 3: fair use exemption for parody; Option 4: exemption for UGC. He is advocating for option 4. Some political uses aren’t parodies in the traditional sense. Hitler Downfallvideos can be made about what’s going on in Hong Kong. Incidental uses, covers on YouTube.
Transplant Canadian model without much change. Transplants have benefits and drawbacks. DMCA is 16 years old—not a great candidate for transplant, known unintended consequences, product of specific US industry bargains. UGC exemption: not based on industry bargains; HK and Canada face similar challenges.
Use must be solely noncommercial; source must be mentioned if reasonable to do so (moral rights respecting); individual must have had reasonable grounds to believe that the copy s/he used wasn’t infringing. Finally, no substantial adverse effect, financial or otherwise, on copyright owner.
His only proposed change: remove “solely” because some internet platforms provide ad revenue. His proposal: “predominantly,” so that would cover people with tiny amounts of ad revenue.
Industry reaction: fails the 3-step test/TRIPS, including TRIPS criminal enforcement provision; also that there’s not much case law on Canada’s exemption; complications on rights in underlying works; loophole for commercial pirates.
IFPI offered counterproposal for parody exemption (not satire, not UGC) when the parody satisfies 9 different conditions.
December, gov’t released discussion paper. Canada is the only country with a UGC exemption (which isn’t right if you also count transformative use). Worried that might fail 3-step test. Many experts agree that Canadian exemption doesn’t comply with the 3-step test.
If you’re really worried about that, can build in a 3-step test to the exemption, as Korea did. Also, officials who say this don’t want to answer the Q whether US fair use violates the 3-step test. If worried, have fair dealing for UGC. Finally, we can have a reciprocal license—individuals can use the content, but allow the authors/© holders to use the new work for noncommercial purposes in return. If concern is about “predominantly,” can redefine noncommercial to explain how to treat revenue that’s covered or not covered—more than trivial economic prejudice, a standard used in other elements of the criminal copyright law. Final proposal: profit-sharing arrangement for commercial UCG, as done with Content ID. Levy system also worth considering.
Could also consider a sunset period as licensing systems change. “Licenses for Europe”—trying to develop microlicenses. If those eventually develop, it will be easier to license, but those don’t exist now. Finally, should be a moratorium on civil/criminal actions against users, as in AHRA in US. “No action may be brought under this title…” Break the interest into two; if the concern is really about intermediaries making money, then it would be much better to protect users from being sued while allowing actions against intermediaries/safe harbors for intermediaries.
Q: what can we do?
A: Anyone at INTA should help explain what’s going on abroad. Foreign professors can be very persuasive.
Q: proposal makes a lot of sense—what are the politics? Do you need compromise/fallback? YouTube could have a default rule that remixes aren’t monetized. These platforms could set up distinctions.
A: fallback at this moment is fair dealing for parody and satire, broadly defined. But not as appealing as what you mentioned. YT is playing both sides—negotiating with IFPI and pushing for reform. Right now YT has master licenses; you’re proposing a noncommercial category, and YT isn’t willing to do that at a global level. At the moment, if they can get a UGC exception, they might still want a master license to deal with commercial exploitation.
Q: (1) Noncommercial use is a difficult concept—people who hope to be ‘discovered’—noncommercial? YT makes money from videos—how to draw the line? (2) Reason to know the copy is infringing—isn’t that circular? (3) Market impact is like factor 4 in fair use—how will it be defined? Isn’t that also circular?
Q: as practical matter, how do you test noncommerciality? Video can be on YT for years w/out making money, then make money, though that’s not usual.
Q: economic impact—really biting parody might harm the market, but we understand that’s not what the market harm factor is about in US. In HK, is that still true?
RT: Noncommercial isn’t difficult. The fact that someone made money when you bought your computer doesn’t make what you write on it commercial—you didn’t make money from that. The fact that you might someday have commercial success doesn’t make everything you do until that point commercial. If either thing were true (third party profit or potential future money) then there would be no such thing as noncommercial use. There is; we’ve lived with it for a while; it is perfectly manageable even if there are some cases at the boundaries, as there are for any definition.
A: noncommercial at the point of posting is his definition. If you suddenly start making money, could reconsider. (Points out that CC license users do have disagreements about boundary cases.)
Biggest problem is that users can’t compensate license holders enough for their demands/ad revenue isn’t enough even if YT is willing to allocate all the ad revenues to rightsholders. Tushnet’s comments touch on whether third parties matter to rightsholders. Thus his proposal to cut the issues in half: moratorium on actions against users while allowing actions against intermediaries. But it’s true that third parties always benefit from fair dealing.