IPSC: TM II/(c) II

Trademark II

Mary Catherine Amerine, The George Washington University Law School, Mind the Gap: How Brand Gimmicks Have Made Infinite the Zone of Expansion

Doritos x Empirical alcohol: on shelves for three years. Tesla mezcal. Le Moutarde Vin wine cobranded with Grey Poupon. Panera Croissant clutch. KFC nail polish.

Cobranding: often well within the wheelhouse of one participant, as in Adidas x Arizona Ice Tea sneakers. But also in house production: In and Out sneaker (after they sued Puma for sneaker design).

If we’re taking this as having an impact on consumers, consumers might expect a lot more product expansion. But we shouldn’t expand “bridging the gap” infinitely for brands that wouldn’t otherwise qualify for famous mark protection. So how should we think about these expansions? Haven’t seen it played out yet, but may be coming.

Sliding scale of similarity? If not famous, maybe only really expansive copying is covered when there is extreme product difference.

Q: could you change the standard so that if P didn’t do a shoe deal it shouldn’t have rights in shoes? Barriers to entry for expansions are so low; maybe we should rely less on hypotheticals and more on what you’ve actually done.

Q: is this unchecked licensing really interfering with the actual TM function of the mark? Not naked licensing but may have similar effects.

Linford: would drive big brands to engage in a lot of branding/licensing deals.

Alex Roberts: many of these were limited editions; maybe we should have abandonment-like theories—abandoned the expansion if not the core brand.

RT: The greater similarity standard would make courts nervous framed that way, but you could do it a different way: these collaborations/licensing deals have shared characteristics. They use the brand prominently and clearly. Not a red wax seal on tequila and no other signifier. So explaining what precisely consumers have been trained to expect—and we do expect consumers to be brand savvy—could help explain what is nonconfusing.

Copyright II

Ari Lipsitz, Boston University School of Law, Tarot's Uneasy Copyright

Rider Waite deck: Public domain under 1909 Act; in 60s, US Games Systems, Inc. came in—wanted to call it the only authorized deck even though in public domain, and foreign publisher said yes. Became a best-seller. If the deck started & remains in the public domain, why does US Games have market dominance?

Strategies: strategic © registrations—registration claims only minor variations but enables © notice on every card. 1971 registration claims added matter; very thin but very hard to figure out what they do and don’t claim—identified the designs as preexisting material. Original registration identified them as being in public domain; revised description 2003. New color schemes, slightly cleaner images—not clear that those really qualify for new ©.

Exclusive licenses: agree not to dispute ©. C&D letters focus on TM and lead to settlements instead of litigation. Permissions: fees for commercial reuse; attribution only for tarot reading.

Evergreening: routinely republishing the deck with minor and significant variations.

Blurring © and TM: tell people that Rider-Waite is a TM owned by them including the images—just like you can’t copy “Monopoly.” [Anti-Monopoly would like a word!] Also claim ownership of artist Pamela Coleman Smith’s name. [Dastar, Dastar, Dastar.]

James Grimmelmann: complex structured set of symbols and meaning occupies an interesting place in the idea/scenes a faire world. Would be great to explain how this is similar to or different from other publishers who retake public domain materials, like printing public domain books.

A: sheet music too!

Q: a story about transfer of wealth and control from authors to publishers. Mark Twain’s complaints about the existence of the public domain were similar.

Bruce Boyden, Marquette University Law School, Levels of Expression

Is selecting which result from an AI prompt is the right one enough to create copyrightability? This is a deceptively hard problem.

Is it creative enough? What constitutes creative enough is basically undefined. We look for evidence of proxies for creativity: choice/judgment/opinion. The more you’re making choices—choosing from 2 is not maybe significant, but choosing 1 from 100 could be?

Determination of creativity is supposed to be holistic—how do we apply that to selection of an image? It’s not selection in the sense of a compilation. You have to refer to what’s being selected to understand what the expression is. Selecting an image to serve as your icon serves an expressive purpose in a 1A sense if not a © sense. Individual components of works are unprotected but can achieve protection as strung together by author. Literary works example indicates important distinction: if you’re only talking about selection of an image, that’s not arranged in any way.

103(b) would probably doom attempts to claim a single output. But are images really preexisting if they don’t preexist the author’s interaction with the AI?

Does the size matter? Are these functionally “microworks”? Hughes writes about smaller parts of a larger project, but he explicitly says that small things like haiku should qualify for ©.

If selection shows the most power/effect on perception when it is a single choice, that suggests that selection alone can’t do the work. Coordination and arrangement are key. [But see databases?] Selection seems more like a starting point but doesn’t define a completed work. It has to be laid out in some sort of order to convey aesthetic or informational expression. Otherwise it’s not a compilation, but just a heap.

Rosenblatt: Selection coordination and arrangement aren’t rules in themselves, they’re proxies for 102/103. So the question we’re asking isn’t whether selection itself is good enough, but whether selection can ever create anything that is nontrivially distinct from an otherwise unownable work. Usually not, in her view, but maybe in a very rare circumstance. Selection of a set of pixels from a larger work of art to create a closeup. Nontrivially distinct from what came before. Mere selection w/o coordination and arrangement might do it, but we’re not asking whether selection is enough, we’re asking “when is selection alone enough” and the answer is “rarely.”

RT: (1) Why are you setting authorship aside? If I go down to the seashore and choose the most beautiful rock, it doesn’t matter how much judgment I exercised. Authoring a choice and authoring a work are not the same thing. (2) Wittgenstein’s definition of games—no necessary and sufficient condition; maybe the same thing is true of copyrightable works when it comes to the limit cases where the creativity comes from the compilation. [Relatedly: Selection alone never enough is the claim: But see databases? Would a poetry compilation be infringed by the exact same selection in a different order (assuming that the selection principle is not, e.g., “these are all the poems of Edna St. Vincent Millay”)? I think the answer is likely to be yes.] (3) Preexisting: the elements need to preexist the creation of the work, not the interaction w/the AI. But courts have always ignored this requirement, so it might not matter.

Grimmelmann: you’re appealing to intuitions about how much information is in a selection. Sarah Scheffler et al.’s paper on using complexity to analyze copyright.