This page collects occasional outside-of-class questions from students in the Spring 2026 edition of Copyright Law, and my responses to them. They are posted here for the benefit of the entire class.
Question(s):
On page 15 of the opinion, the Court summarized, “In sum, the first fair use factor considers whether the use of a copyrighted work has a further purpose or different character, which is a matter of degree, and the degree of difference must be balanced against the commercial nature of the use. If an original work and a secondary use share the same or highly similar purposes, and the secondary use is of a commercial nature, the first factor is likely to weigh against fair use, absent some other justification for copying.”
It’s the second sentence requiring “some other justification for copying” when a similar secondary work is of commercial nature that made me think that commerciality does create a sort of a presumption against fair use.
Relatedly, in Sandra’s hypo, I thought that the secondary ChatGPT use of the Madonna photo shared too similar a purpose (and character, according to the Warhol opinion), as it licensed the Madonna photo. This essentially replaced the WB license (you also mentioned Warhol, but how would they make a claim? Would copyright be based on style or collection of art?). Because the ChatGPT use is commercial in nature, it needs “some other justification for copying.” I assume that the lack of accessibility of getting permission would be relevant here (though, would it be that hard to find a WB representative that deals with copyrighted works since they are such a big corporation?).
As for damages, WB would go after Sandra, much like how Goldsmith went after AWF and not Vanity Fair (or its parent company). Would WB go after the coffee shop too? In other words, how far does liability extend? And could the coffee shop owner make a defense as a good faith buyer?
Response:
As always, I welcome the questions!
On the first – “some other justification for copying” – I read that as a reference to 102(b) and related distinctions between what a copyright covers and what a copyright does not cover (ideas v expression, facts v expression, material object v intangible “work of authorship”) (in this course, that material comes relatively soon), power to copy that comes with joint authorship, and to compulsory and statutory licensing (that comes toward the end).
On the second – I mentioned Warhol because I do think that there is a hypothetical claim that the AWF might make. They should lose that claim (the usual rule, as you note, is “no copyright in style”), but it’s worth considering. Whether ChatGPT is “commercial” for all purposes is an open question. Yes, OpenAI is a big company that makes a lot of money via its products, but one of the lessons of the Warhol opinion is that one has to measure “commerciality” against the specific thing that the defendant is accused of doing that violates the Copyright Act. What, precisely, did OpenAI do here (in this imaginary case) that would lead to a finding of copyright infringement, absent fair use? As for WB, the usual practical rule for movie studios and producers is: All Your Work Are Belong To Us. Yes, at least nominally, there is a way to send a permissions request to WB. Practically speaking, the ability to get someone at WB to pay attention and be reasonable is limited. WB and its ilk want to have their cake and eat it, too. “Yes, we grant permission,” but “in practice, that never happens, or the price is extremely high, or reasonable negotiations are limited to our commercial partners.” The real world perspective is that WB etc. are in the business of making money, not in the business of catering to every little person who wants a “yes” answer.
As for extensions of liability, the coffee shop is mostly in the crosshairs, rather than Sandra, because it’s the coffee shop that is (hypothetically) displaying the work in a commercial setting (that’s covered by the exclusive rights in section 106). “Good faith buyer” (or “purchaser”) is a concept that comes from commercial law, which has no relevance here. If the work as used infringes (absent fair use), then it infringes. This is why art purchasers routinely include indemnity clauses in contracts with dealers and distributors, and why dealers and distributors routinely resist agreeing to them – unless the dealer or distributor can find relevant insurance (which they can do).
The dialogue continued:
Follow up to my follow up: I’m not understanding how referencing 102(b) changes the holding or analysis? Could you elaborate?
Response:
To your question about 102(b):
You quoted the Court in Warhol: “In sum, the first fair use factor considers whether the use of a copyrighted work has a further purpose or different character, which is a matter of degree, and the degree of difference must be balanced against the commercial nature of the use. If an original work and a secondary use share the same or highly similar purposes, and the secondary use is of a commercial nature, the first factor is likely to weigh against fair use, absent some other justification for copying.”
I mentioned 102(b) in connection with addressing the last phrase: the four fair use factors taken together weigh in this direction or that direction “absent some other justification for copying.” What might that “other justification” be? 102(b) is the first and most obvious answer to that question (though not the only one). Any copyright has to be dis-assembled, in practice, to sort protectable 102(a) “expression” from 102(b) “idea, procedure, process, system, method of operation, concept, principle, or discovery.” Copying the latter is always permitted, even if that material is copied from an otherwise copyrightable work of authorship. Lynn Goldsmith has a copyright in her photo of Prince. She does not own a copyright in “a photo of Prince.” Her copyright does not extend to other photos of Prince unless those photos appropriate her specific expression.
Further question:
It seems that the Court provides an example of “other justification” in footnote 8 of the sentence we’re examining (or alternatively, the footnote highlights the sufficient difference in use). It doesn’t explicitly mention 102(b), but I wonder if Google’s use of Sun’s code could qualify under 102(b). I might read the Google case to get a better understanding here.
Yet another response:
Footnote 8 is a bit convoluted. As I read it, the Court is saying that Google’s copying of Oracle’s code was justified in part because of the functional attributes of the code. (The “coders need to be able to code” argument doesn’t make a lot of sense to me, but we might come back to that some other time.) The Court tries to re-write a functionality argument in terms of “new systems for new products,” but “new systems for new products” is a way of saying that the code makes different computers operate, not that the “expression” in the code is somehow different in different settings. In a word, the code is functional. If that’s right (perhaps you or others might come up with a different interpretation?), then it’s really 102(b) sneaking into 107.
Read Oracle v. Google, then read backwards through precedent to Atari v Nintendo in the Federal Circuit and Sega v Accolade in the Ninth Circuit, both from 1992, if I recall correctly. These are the foundational cases having to do with reverse engineering, and whether copying code in order to reverse engineer how it works constitutes fair use. The short version of the holding in both cases (though the facts in each one differ somewhat) is that copying a copyrighted computer program is fair use so long as and to the extent that the purpose of the copying (factor 1) is to obtain information about the functionality of the code, which is unprotected under 102(b).
Why, you might ask, didn’t the Supreme Court in AWF footnote 8 simply cite to Sega and to Atari rather than making essentially the same argument in a more oblique way? I don’t really know, but I do know that as a matter of practice, the Supreme Court strongly resists citing lower courts ever, for anything. The Court often behaves as if the only law that exists, and the only law that matters, is the Court’s own precedents.
One more question:
First, the Court in Google bypassed the analysis of whether Sun’s code was copyrightable (presuming it was) and jumped into a fair use analysis. Justice Thomas, in his dissent, vehemently disagreed with this because the majority ultimately distinguished between a functional code (“declaring code”) and expressive code (“implementing code”), creating a 102(b) gloss (Google argued this was squarely a 102(b) case), which permeated their fair use analysis. My take on this is that Justice Thomas felt the majority’s reasoning was distorted because had the majority analyzed the copyrightability of Sun’s code, then the code would be uncopyrightable from the start due to its functional nature.
Second, the whole discussion around function and expression came during the “nature of the copyrighted work” portion of the opinion (see quote below; bolded for emphasis). Not sure if this is significant, but the majority tackled this issue first, even though it’s the second prong of the analysis, which Justice Thomas took issue with because he believed precedent demonstrates that there is a hierarchy in fair use factors.
These features mean that, as part of a user interface, the declaring code differs to some degree from the mine run of computer programs. Like other computer programs, it is functional in nature. But unlike many other programs, its use is inherently bound together with uncopyrightable ideas (general task division and organization) and new creative expression (Android’s implementing code). Unlike many other programs, its value in significant part derives from the value that those who do not hold copyrights, namely, computer programmers, [*29] invest of their own time and effort to learn the API’s system. And [****43] unlike many other programs, its value lies in its efforts to encourage programmers to learn and to use that system so that they will use (and continue to use) Sun-related implementing programs that Google did not copy.
Google LLC v. Oracle Am., Inc., 593 U.S. 1, 28-29, 141 S. Ct. 1183, 1202 (2021)
Lastly, my interpretation of “other justification” in footnote 8 of the Warhol opinion is the justification to promote the creativity of programmers to make new programs and apps for a mobile platform. Footnote 8 cites from the “purpose and character” portion of the Google opinion, which interestingly does not mention the function v. expression dichotomy. The “purpose and character” section of the Google opinion only mentions the context of why Google needed to copy a specific portion of code for its new product.
Response:
Reading Supreme Court opinions in IP cases is like reading law through thick gauze.
The cases have been intensely lawyered through lower courts, often for years and years, and the Court is inundated with amicus briefs on all sides and often (also) has a Solicitor General’s brief as well. That’s just the input. The Justices themselves rarely have any authentic understanding of IP law. There is no one on the Supreme Court today who had any meaningful engagement with IP law as a lower court judge or practitioner. (Sotomayor probably comes closest.) In recent years, Justice Breyer knew his way around IP law. Justice Ginsburg had a good guide, behind the scenes, because her daughter is a sophisticated copyright scholar at Columbia Law School.
That’s just for starters. More: the Court is usually at least vaguely aware, especially in high tech cases, that what it might say in one area of IP (say, copyright) is going to have echoing effects in others (say, patent, and the reverse). One reason to steer clear of “copyrightability” issues in Google is to avoid getting tangled up in software patent law coming out of the Federal Circuit, which has been a mess. Fair use in copyright is, in a way, its own reasonably self-contained silo. But courts and litigants at times tiptoe around “settled fair use understandings” rather than diving into new analyses of the law, because “settled fair use understandings” end up getting baked into both industry and consumer practices. Later in the semester, we will come across cases that might have been litigated as fair use cases but where both sides agreed not to argue about fair use in order not to disrupt conventional fair use understandings in that context.
Then, when you get to the details of the factor-by-factor analysis: Justice Thomas is quite wrong about there being a “hierarchy” in fair use. Precedent says that the first and fourth factors usually weigh the heaviest, but that does not mean that a court has to start with the first one or that the second and third might not influence what a court does with the first and fourth. It’s a “holistic” analysis. The “scope” of purpose and character – does it go narrowly to this work and this use, or does it go more broadly to sectors or industries? – is an unresolved question. And the emphasis on “what future programmers might need to do” is, I think, just odd. There is a robust tradition in copyright, going back especially to the Feist case from 1991 (we will see this soon) that says: labor and effort interests of human beings do not count in copyright analysis; what counts is the work itself – the output. “Progress” is examined from the standpoint of: does society get more and better things to read (listen to, watch, etc.)? Also as we will see, despite that precedent being quite clear, lawyers continue to lean into that style of argumentation rhetorically, because unsophisticated courts (here, the Supreme Court) pick up on it, and juries do, too.
In sum: I’m glad to read how deeply you’re investing in this topic. Part of learning to read and “do” copyright includes acquiring a lot of context that allows you to read skeptically. The more cases you read, the more sophisticated your analysis will get.
Question:
From my reading, it seems like the Halicki court contradicted the Towle court by focusing on anthropormophism as a key consideration for conceptual quality (prong 1 of the Towle test), when the Towle court made no such analysis. In fact, it seems like the Halicki court cherry picked the Towle court’s creative description of the batmobile’s crime fighting and techy characteristics (to support the distinct characteristics in the second prong) as a way to support the anthropomorhpic nature of the batmobile.
In my mind, “sentience” is a concept that doesn’t seem to be supported by Towle but the Halicki court nonetheless pivoted to anyways. However, I understand the logical appeal behind adding sentience as an element to “character” that deserves copyright protection. Let me know what you think.
Response:
I appreciate your close reading.
My thoughts, which we’ll come back to in Wednesday’s class when we lean back and look more generally at industries and sectors:
To me, doctrinally, the Towle case is just nuts. I can, barely, get behind the instinct that fictional characters are copyrightable independently of the larger works (series of works) in which they are embedded. James Bond. Sherlock Holmes. I really cannot get behind the idea that a car is a character rather than a material object. But doctrine was being used here for strategic purposes; I don’t like the doctrine, but I understand and have at least a modicum of sympathy for the strategy. DC, its allies in the case, and Hollywood generally were using the case partly to get at the proposition that extracting any money from any of its “properties” without permission is not OK, and partly to get that strategy documented in a Ninth Circuit opinion. Why? The revenue model of a modern Hollywood movie depends on a lot more than ticket sales and other distribution revenue. (The founder of that model is George Lucas.) “What is the thing?” was being manipulated, given the fluidity in the doctrine, for what are (to me) obviously cynical but comprehensible ends. If we like movies, then the law has to be at least somewhat sympathetic to how movies are made. And financed.
(If we back up from the details of Towle and look at the doctrinal question less from a “does business need this?” standpoint and more from a “is this consistent with the general scheme of copyright?” standpoint, the answer seems (again, to me) to be clearly: neither DC nor its allies need copyright protection for characters generally or for “prop characters” specifically in order to underwrite the incentive to manufacture creative things like films and TV shows. If there is no plausible incentive function, then there is no need for the copyright. If you talk to folks in Hollywood, they’ll disagree on the broad point (many Hollywood execs will say: copyright in characters is fundamental to how modern Hollywood works) but might broadly agree with the narrow one (it’s difficult to say “the Batmobile is a character” with a straight face, even if we all understand the imperative to make money any way they can).)
Now to Halicki:
I think that the disconnect that you’re reading is evidence of the court’s discomfort with the broad reading of Towle and, by extension, discomfort with a broad reading of the kinds of merchandising rights that DC and its Hollywood allies are using the Towle litigation to try to control.
Copyright is a system of limited rights; just because fans love the “thing” that the copyright owner controls does not mean that the same copyright owner is entitled to collect every last penny of economic value associated with that affection. Fans get to be fans; fans get to participate in an economically- and financially-grounded version of fandom. (Even Hollywood understands that, at some level.) Doctrinally, eventually, Towle really does run into the idea/expression dichotomy, into fair use, and into the proposition that so-called “PGS works” (see class 11 on the syllabus) cannot be copyrighted without a deep separation of material v immaterial dimensions).
For a variety of reasons, lots of them (I think) having to do with how the current Halicki litigation was framed (breach of an earlier settlement agreement, among other things), the case didn’t tee up those issue. The defendants in the Towle case didn’t pitch hard on those issues. Courts pick up the doctrinal themes that the parties argue (sometimes) and the courts think are the easiest ways to get out of Dodge (sometimes).
Here, the court could have said: this is an idea / expression case rather than a “sentience” re-reading of Towle. Or it could have said: there is no “originality” embedded in Eleanor, as there is (arguably) in the Batmobile. It could have said: this is a possible PGS work (again, class 11, which we haven’t come to yet), but “it’s just a car” with no copyrightable originality that is separable from its materiality. Instead, the court said: we’ll interpret Towle narrowly. Implicitly, maybe the court is saying that the Batmobile is a unique “character” in the history of film automobiles? (I can think of other “sentient” cars – Herbie, the Love Bug, for one, and KITT, in Night Rider, for a second – but robots are the biggest class of possibly relevant “things.”) Animated characters aren’t problems (the Cars movie, for example), because an animated car is still an animated image and therefore easily treated as a “work of authorship” from a visual art standpoint.)
Does that make sense?
The takeaway remains: the “thingness” of a copyrightable work of authorship is a squishy and manipulable concept. Character copyrights are simply an interesting and fun way to introduce the problem.