How a Mismatch Between the House Report and the Enacted Section 301 Corrupted Copyright Preemption Doctrine

Thursday, April 16, 2026 

Executive Summary

Instructor(s)
R. Anthony Reese, Professor of Law, UC Irvine School of Law
Marketa Trimble, Professor of Law, University of Nevada Las Vegas William S. Boyd School of Law
Guy Rub, Professor of Law, Temple University Beasley School of Law
Shyamkrishna Balganesh, Professor of Law, Columbia Law School
Peter Menell, Professor and Faculty Co-Director, Berkeley Center for Law and Technology, UC Berkeley School of Law

Keywords

Section 301 Copyright Act • preemption equivalency test • contract • breach of contract • hot news • misappropriation doctrine • International News Service v. Associated Press • preemption • NBA v. Motorola • misappropriation • legislative history error • Second Circuit • Genius Media Group v. Google • breach of contract • preemption • data scraping • X Corp. v. Bright Data • conflict preemption • obstacle preemption • fair use • federalization • unpublished works • 1976 Copyright Act • common law • copyright national uniformity • state copyright statutes • Georgia v. Public.Resource.Org • government edicts • pre-1972 sound recordings • Section 1401 • state law ownership • choice of law, “is breach of contract preempted by federal copyright law Section 301,” • “does the hot news misappropriation doctrine survive copyright preemption after NBA v. Motorola” • ProCD v. Zeidenberg • extra element test • promise consideration • copyright policy • AI training

Legal Analysis

Federalizing Unpublished Works: How Publication Ceased to Divide State from Federal Copyright Protection

The decision embodied in the Copyright Act of 1976, Pub. L. No. 94-553, 90 Stat. 2541, to bring all unpublished works under federal statutory protection from the moment of creation was, as Professor R. Anthony Reese demonstrated, far more contested and doctrinally consequential than its subsequent treatment as a settled background principle would suggest. Register Abraham Kaminstein himself described the question in 1965 as “perhaps the most fundamental issue underlying the entire revision program,” and Barbara Ringer characterized it in 1961 as “the pivot point on which the whole revision program turns.” Reese challenged the standard account of a clean state-to-federal transition keyed to publication, documenting that the Copyright Act of 1909, Pub. L. No. 60-349, 35 Stat. 1075, under Sections 11 and 12, already permitted voluntary federal registration of unpublished works in eight enumerated classes—lectures, dramatic and musical compositions, photographs, works of art, plastic works, technical drawings, motion pictures, and motion picture photoplays—with the result that, from 1929 to 1977, 25 percent of all non-renewal Copyright Office registrations were for unpublished works, and among the eligible classes, over 75 percent of registrations were for unpublished works; for musical compositions alone, 83 percent of registrations were for unpublished works. The Copyright Office’s 1961 registers report initially resisted comprehensive federalization, recommending instead that “public dissemination”—communicating a work to the public by any method—replace publication as the dividing line, in order to preserve common law protection for genuinely private works; the report stated that “a fixed time limit for the protection of private papers against unauthorized disclosure would be undesirable” and that “the right of privacy should be paramount.” That position collapsed within a year under overwhelming opposition that condemned the dissemination standard as “nebulous, confusing, dangerous, amorphous,” and by 1963 Register Kaminstein announced that “the advantages in a single federal system outweigh the advantages of preserving common law copyright for undisseminated works,” a complete reversal driven primarily by three concerns: the unfairness of registering only certain classes of unpublished works, the chaotic uncertainty of publication doctrine—Kaminstein describing it as “chaotic and becoming worse all the time”—and the risk that works commercially exploited by broadcast or phonorecord distribution might enjoy perpetual state common law protection while avoiding every statutory limitation. The move to a unitary system had ripple effects throughout the revision bill, most significantly creating what Reese has termed the “unpublished public domain”: for the first time in American copyright history, unpublished works became subject to finite duration, so that in 2003 every unpublished work by a person who died before 1933 entered the public domain simultaneously—what Reese characterized as “probably the largest single deposit ever into the public domain.” The anomalous exclusion of pre-1972 sound recordings from the federal system was, Reese argued, a direct consequence of the unresolved state-law question whether phonorecord distribution constituted publication, a question that generated “a significant furor in the copyright bar” in the 1950s and was not finally resolved until Congress enacted 17 U.S.C. § 1401 in 2018, which Reese characterized as “the final removal of the last bit of rubble from the 1976 Act’s destruction of the wall between federal and state copyright protection.”

State Copyright Statutes, Section 1401’s Choice-of-Law Gap, and the New Frontier of AI Transparency Legislation

Despite the 1976 Act’s broad preemptive sweep, a residual body of state copyright-specific legislation persists—and is growing—in ways that Professor Marketa Trimble characterized as simultaneously outdated, constitutionally precarious, and occasionally innovative. The threshold problem is widespread obsolescence: most state statutes still use formalistic language directing state agencies to “copyright,” “cause to be copyrighted,” or “secure copyright” in specified works, treating registration as constitutive of protection in terms that Colorado alone has updated to reflect the 1976 Act’s elimination of registration as a prerequisite to protection. Trimble documented a range of state statutory artifacts—Pennsylvania’s direction to register geological and topological surveys, Arkansas’s authorization to copyright a state history book, Maryland’s provision for official insignia of farm products, Ohio’s and Oregon’s highway maps—as illustrations of “what state legislators think is worth copyrighting,” many of which sit in direct tension with Georgia v. Public.Resource.Org, 590 U.S. 255 (2020), which held that not only judicial decisions but also legislative materials may be unprotectable under the government edicts doctrine. On the affirmative side, Trimble identified Section 1401(l)(2) as creating an unresolved structural deficiency: the provision defers to the law of “any state” to identify who holds the exclusive right to reproduce a pre-1972 sound recording, thereby defining the rights-owner under the new federal regime by reference to a body of state law that varies across jurisdictions and for which the federal act provides no choice-of-law mechanism. Trimble cautioned that states such as West Virginia follow different ownership rules, creating “an interesting situation where we could have multiple, apparently, owners of rights under Section 1401, with no choice of law provision in the federal act.” On the innovative frontier, Trimble surveyed a wave of state legislation addressing AI-adjacent concerns: Arkansas has enacted a statute addressing ownership of content generated by and of models trained on AI tools; New York has introduced a bill imposing disclosure obligations on crawlers serving journalism providers; and California’s AI training transparency statute represents an area of active state experimentation that could face preemption scrutiny. She also identified Connecticut’s digital e-book lending statute—which does not restrict the copyright owner’s rights directly but limits libraries’ contractual freedom to accept restrictive publisher license terms—as a more legally durable approach than the Maryland digital lending bill that courts struck down, illustrating how states are learning to calibrate their interventions to avoid express preemption under Section 301.

Section 301’s Equivalency Test, the ProCD Majority Rule, and the Case for Conflict Preemption in an Age of Data Scraping

The preemption of state contract claims under Section 301 of the 1976 Act presents what Professor Guy Rub described as an analytically broken framework that asks the wrong question and produces results that are simultaneously “under exclusive and over exclusive.” Section 301(a) expressly preempts only state causes of action “equivalent to” the exclusive rights enumerated in Section 106—a standard the House Report predicted would produce clarity so self-evident as to “foreclose any conceivable misrepresentation” and avoid “the development of any vague borderline area between state and federal protection.” As Rub observed, “oh boy. It is extremely, extremely complicated.” The approximately 350 litigated opinions applying Section 301 to contract claims have produced two principal approaches. The minority rule holds that a contractual promise to refrain from any act listed in Section 106—reproduction, adaptation, distribution, public performance, or public display—is itself preempted regardless of the contractual context. The majority rule, indelibly associated with ProCD, Inc. v. Zeidenberg, 86 F.3d 1447 (7th Cir. 1996), treats the contractual promise or consideration as the “extra element” that distinguishes a contract claim from a copyright claim, with the result that virtually every contract action escapes preemption because every contract contains a promise; Rub identified ProCD‘s dominance as attributable to two factors—it is easy to apply as a bright-line rule, and the minority’s fact-specific approach has produced opinions that are “extremely inconsistent” and “messy.” The Second Circuit’s recent holding in Genius Media Group, Inc. v. Google LLC applied this framework to find preempted a breach-of-contract claim arising from Google’s scraping of song lyrics from the Genius website, reasoning that the terms of service merely regulated the public display of copyrighted material; the Solicitor General’s amicus at the certiorari stage implicitly signaled that browser-app-based restrictions should be treated differently from other contractual limitations—a distinction Rub characterized as practically important given the volume of non-copyright-protected factual content on commercial websites. Rub argued that Section 301’s equivalency framework structurally fails because “contract can create a lot of harm to our copyright policy without trying to replace it”—contracts can restrict fair use, extend de facto protection beyond copyright’s scope, and foreclose uses of unprotected information, none of which the equivalency test reliably catches. He endorsed the conflict preemption framework developed by Judge Leval in In re Jackson and applied by Judge Alsup in X Corp. v. Bright Data—which preempted Twitter’s data-scraping prohibition on the ground that the platform could not contractually extinguish users’ potential fair use claims—arguing that this framework allows courts to ask the normatively relevant questions the equivalency test forecloses: “Was it a negotiated contract or not? Is it content trying to promote privacy or other interests or commercialization? What market power do the parties have?” Professor Shyamkrishna Balganesh then supplied the historical foundation for understanding why Section 301 is structurally ill-suited to these questions, reconstructing the corrupted legislative history of the hot news misappropriation doctrine within the preemption provision. The hot news doctrine originates in International News Service v. Associated Press, 248 U.S. 215 (1918), where the Supreme Court under Justice Pitney’s majority opinion invoked equitable jurisdiction and a theory of quasi-property to enjoin time-sensitive free-riding on AP’s news-gathering; the doctrine survived Erie Railroad Co. v. Tompkins, 304 U.S. 64 (1938)’s elimination of federal general common law by migrating into state common law, though its scope remained contested. In the Section 301 drafting process, misappropriation was deliberately excluded from the 1965 bill—the Register’s Supplementary Report characterizing it as “the virtual equivalent of copyright”—then reinserted by a Department of Commerce amendment in the 1975 hearings, and finally eliminated from the enacted statute after the Department of Justice sent a last-minute letter to Senator Hugh Scott’s office condemning the doctrine as creating monopolies, raising unexamined antitrust concerns, and threatening to “effectively neutralize the logic behind preemption.” The critical error is that the House Report was never updated to reflect the elimination: it still contains an elaborate endorsement of misappropriation’s “structural nonequivalence” to copyright and its suitability for preserving state equitable flexibility—language that, as Balganesh demonstrated, “is contingent on there having been misappropriation referenced in the statutory provision, which never makes its way into the final bill.” The Second Circuit’s 1997 decision in National Basketball Ass’n v. Motorola, Inc., 105 F.3d 841 (2d Cir. 1997), relied on precisely that obsolete House Report to resurrect misappropriation as a robust doctrine, and the error propagated from New York to Ohio, Missouri, Illinois, and California before the Second Circuit partially corrected it in 2011. Balganesh concluded that the “origins of this comes from a fundamental error in the House report and its belief that misappropriation was somehow wholeheartedly endorsed, when, in fact, the opposite was indeed true”—leaving a body of circuit and district court doctrine built on a report that describes a statute that does not exist, with live consequences for AI training data disputes and news aggregation preemption claims moving through the courts today.

Generated by AI based on the Interview/Transcript below.

Key Takeaways

• The House Report’s misappropriation endorsement has no statutory basis. Balganesh demonstrated that the House Report’s elaborate discussion of misappropriation’s “structural nonequivalence” to copyright describes a version of Section 301(b)(3) that was eliminated from the enacted statute, meaning that courts citing that legislative history to resurrect the hot news doctrine have been “relying on the wrong legislative history” for decades.
• Section 301’s equivalency test asks the wrong question. Rub argued that the framework systematically fails to identify contracts that damage copyright policy without replicating copyright’s exclusive rights, and that conflict preemption—asking whether the contract is an obstacle to federal copyright policy—is “a better framework” that “allows us to ask questions that are more relevant than the formalistic is it equivalent or not.”
• The federalization of unpublished works was fiercely contested, not consensual. Reese documented that the Copyright Office reversed its position entirely between 1961 and 1963, abandoning the “public dissemination” dividing line after stakeholders condemned it as “nebulous, confusing, dangerous, amorphous,” and that the change was driven by the need to eliminate perpetual state protection for commercially exploited but technically unpublished works.
• Section 1401’s deference to state law creates an unresolved ownership gap. Trimble identified a structural deficiency in 17 U.S.C. § 1401(l)(2): the provision defines rights-ownership in pre-1972 sound recordings by reference to state law but provides no choice-of-law rule, creating the possibility of “multiple, apparently, owners of rights under Section 1401” in states with divergent ownership rules.
• ProCD’s majority status reflects administrative convenience, not doctrinal soundness. Rub observed that ProCD‘s dominance across most appellate circuits is attributable to the fact that it is “easy to apply” as a bright-line rule, while the minority’s fact-specific approach has produced opinions that are “extremely inconsistent” and “messy”—an asymmetry that has allowed a heavily criticized rule to entrench itself in the law.
• NBA v. Motorola’s misappropriation resurrection was built on error. Balganesh argued that the Second Circuit’s NBA v. Motorola decision was “wrong” as a matter of statutory interpretation, and that the error spread uncorrected for fourteen years across multiple circuits that have not revisited their own reliance on the flawed legislative history.
• Pre-1972 sound recordings were a structural residue of unresolved publication doctrine. Reese argued that the differential exclusion of pre-1972 sound recordings from the 1976 Act’s federal system was a direct consequence of the unresolved question whether phonorecord distribution constituted publication, and that Section 1401’s 2018 enactment represented “the final removal of the last bit of rubble” from the publication wall.
• State AI transparency legislation is the new preemption frontier. Trimble identified a wave of state statutes and bills—Arkansas’s AI-generated content ownership act, New York’s crawler disclosure bill, California’s AI training transparency statute—as occupying contested preemption territory, with the question of whether they address copyright-equivalent rights or supplement copyright policy remaining unresolved.
• Conflict preemption foreshadows AI training data outcomes. Rub predicted that Judge Alsup’s reasoning in X Corp. v. Bright Data—treating terms of service that extinguish fair use claims as conflict-preempted—is “foreshadowing a decision that you will come to a year later with AI data,” suggesting that contractual restrictions on AI training of publicly available content may face conflict preemption challenges on similar grounds.

B-CLE Recording (CLE: FREE) | Youtube Recording | Resource(s) | Speaker Bio(s) & Contact Info

Download the interview/transcript and slides here!

Interview/Transcript

This interview/transcript was based on a conversation on April 16, 2026 about 29th Annual BTLJ-BCLT Spring Symposium: Origins, Evolution, and Possible Futures of the 1976 Copyright Act, hosted by Berkeley Center for Law & Technology, UC Berkeley School of Law. The panel on ‘The Shifting Line Between Federal and State Protection’ was provided by R. Anthony Reese, Professor of Law, UC Irvine School of Law, Marketa Trimble, Professor of Law, University of Nevada Las Vegas William S. Boyd School of Law, Guy Rub, Professor of Law, Temple University Beasley School of Law, Shyamkrishna Balganesh, Professor of Law, Columbia Law School, with closing remarks by Peter Menell, Professor and Faculty Co-Director, Berkeley Center for Law and Technology, UC Berkeley School of Law.

Shyamkrishna Balganesh  00:00

Okay, why don’t we go ahead and get started? Welcome back to our final panel of the day. Our job is to keep you entertained for another hour and a half while nerding out about copyright as we go back to our core topics on the evolution of the 1976 Act, and this one is going to be about federalization and the shifting line between federal law and state protection and the topic of national uniformity. I am joined on the stage today by three very distinguished speakers, Tony Reese from the UCI School of Law, Guy Rub from Temple School of Law, and Marketa Trimble from the University of Nevada, Las Vegas. So Tony Reese will kick things off, followed by Marketa, and then Guy and then I will jump in and talk a little bit as well, after which we will have a conversation about each other’s presentations before opening up to all of you for your most interesting questions. So without further ado, Tony.

Anthony Reese 01:06

Thank you, Shaym, thank you to the organizers for inviting me and to the center and the journal for the terrific event. So my title, or at least my subtitle, is probably a surprise, not a surprise that there was a most fundamental issue in the revision process, but maybe a surprise that it was extending federal copyright to unpublished works. But nonetheless, these are the words, where do I have to point to? There we go. These are the words of Register Abe Kaminstein, I kind of resist calling him Kami, but if you read enough of the panel discussions, everybody calls him that seems like a friend, but I’m going to call him Register Abe Kaminstein, who, in 1965 wrote in the register supplementary report that perhaps the most fundamental issue underlying the entire revision program is whether it should retain the present dual system of common law, copyright for unpublished works and statutory copyright for published works, or adopt a single system of federal statutory copyright from creation. He’s not alone. In 1961 Barbara Ringer said that the question was quote, the pivot point on which the whole revision program turns, prefiguring Jane’s quote this morning about her warning that it was a mistake to view the unified federal system as merely a procedural change. And in a 1977 UCLA Law Review Symposium on the recently enacted, but not yet effective 1976 act, Ralph Brown wrote in A Cheerful Requiem for Common Law Copyright that, quote, The greatest change of the 1976 Copyright Act has been attended with the least controversy. But, I think today the unitary nature of the federal copyright law is simply taken for granted and as a background principle. So I think it’s useful to look at how we got here and why it seemed so significant to the people who wrote the 1976 Act. The standard account of the evolution of the division between state and federal authority and copyright law runs, I think, something like this, from 1790 until 1978 state law protected unpublished works of authorship from the time they were created until the time they were published, and when they were published, state protection ended, and the work either obtained federal protection if formalities were complied with or went into the public domain. By contrast, from 1978 on, all works are protected by federal statutory copyright as soon as they are created, so that unpublished works get their protection under federal law, not state law. This account essentially pictures the landscape before the 1976 act as a wall between federal and cop and state copyright protection, with the act of publication providing the key to unlock the door in the wall and pass through from the state side to the federal side. And instead of asking AI to generate an image of that, I’ll ask you to make your own mental picture of that metaphor and exercise your own imagination, and you’re going to have to make some adjustments to it as we go along. So the first part of that standard account emits some important nuance. In fact, the division between state and federal protection based on publication was more complex before 1978 than the standard account, and was never total. First of all, from 1790 until 1909 federal law provided a cause of action in federal court for unauthorized copying of unpublished manuscripts. There aren’t a lot of cases, but there are some under that federal cause of action. So there was, for a long time, a second door in that wall that didn’t require publication as a key to open, but it was only a procedural door. It got you into federal court for your claim. More significantly, the 1909 Act dropped that federal cause of action, but simultaneously gave federal law a much greater role in providing copyright to unpublished works, because from 1909 to 1978 certain types of unpublished works could obtain federal statutory copyright simply by registration in the copyright office. So this was in, okay, is there any way to go past the first slide? No, I want to go forward. Yes, okay, so what started as section 11 and became section 12 of the Act allowed. There’s animation on the slide. Can I proceed through that? No. Well, okay, yes, so, so under the 1909 act among the classes that you could get protection for were a lecture, next, right, a dramatic musical or dramatic or musical composition, a photograph, a work of art, a plastic work, or a drawing of scientific or technical nature, or a motion picture other than a photoplay or a motion picture photoplay basically categories where works were commonly performed or exhibited before or instead of being published. Now I confess that I’ve generally thought of this provision as essentially a footnote in the standard account, and I think I’m not alone. I’m the co author of a case book that has just a single parenthetical sentence noting it, although that may change in the next edition, but this new option in the 1909 Act turned out to mark a significant shift in the divide between state and federal protection. The best overall picture of this is from the registration data in the Copyright Office and the reports from 1929 to 1977 before 1929 the reports don’t break out published and unpublished. So we have to start in 1929 until I go back through every issue of the catalog of copyright entries to fill in the missing data. And from 1979 to 1997 it turns out 25% of total non renewal registrations were for unpublished works. That is, one quarter of federally copyrighted works were unpublished works. This obviously tells a different story about the divide between state and federal protection. But as substantial as that total number is, it perhaps understates the importance of federal protection for unpublished works, because, of course, registration was only an option in these eight classes out of 13. The Copyright Office annual reports don’t break out the unpublished works by class, but the examining division did produce annual reports from 1954 to 1977 and thanks to Professors Zvi Rosen, those are available for the most part on the Internet Archive, and they have that breakdown. So for those 24 years, turns out over 75% of the total registrations in the eight eligible classes were for unpublished works, and even greater is the significance in some of those eligible classes compared to others. So most notably for musical compositions, which were 28% of all domestic non renewal registrations during this period, 83% of all musical work registrations were for unpublished works. Although this predominance doesn’t develop right away, it took about 20 years from the enactment of the 1909 act for unpublished news of compositions to outnumber annual registrations of published compositions. It was also significant for dramatic compositions, there are many fewer registrations here, 77,000 instead of 1.4 million, but nearly 88% of all registrations for dramatic compositions were for unpublished works. For the other classes, it was less significant, both in terms of the total number of registrations. So this is listing in descending number of unpublished works registered in the period, but in some cases, a pretty substantial percentage. So the 1909 act in Section 11, which became section 12, opened yet another portal in the wall, again, adjust your mental image, between state and federal copyright protection, one that, again, did not require publication as the key to pass through, with the result that substantial numbers of unpublished works crossed over into federal protection. So how then did we come to completely tear down this wall in the 1976 Act? And was that change quote attended with the least controversy, as Ralph Brown said. Well, it began with the revision revision studies that have already been mentioned. One of those studies addressed protection for unpublished works, and articulated three basic concerns. First, that allowing voluntary registration only in these limited classes was unfair, particularly for non dramatic musical works. Apparently, the Copyright Office got lots of requests to register books as unpublished works and had to turn them down. More importantly, there were concerns about the uncertainty of what did and didn’t constitute publication. Again, Register Kaminstein in 1965 wrote, the courts have evolved different concepts of limited and general publication and of investive and divestive publication, all with different consequences, whether publication results from the sale of phonograph records, the exhibition of artwork. So the deposit of copies in a public office is unclear. The President situation is chaotic and becoming worse all the time. So that doesn’t make for a very effective dividing line. And to go back to Bob’s point earlier today about White-Smith and Copies versus phonograph records, in the 1950s judges started to take a different view of that. And there are judicial statements from at least 1950 that judges, these judges thought that distributing phonograph records would constitute publication of the work recorded, the musical composition or the literary work, though there are no actual holdings, but they’re surprisingly adamant in these cases, and the Judge Learned Hand, in a dissent in a 1955 case, took the same position involving what we would now call the sound recording, rather than the composition. And it’s not until 1976 that we get an appellate court decision in the Second Circuit, the Rosette case going the other way. So this injected a lot more uncertainty in the 1950s about publication and created a significant furor in the copyright bar. And perhaps most significantly, the concern was expressed that common law, state, common law protection allowed perpetual potentially perpetual, potentially perpetual protection, as opposed to protection for limited times. And copyright owners could now substantially exploit their works, especially by performance or broadcasting or potentially by distributing phonorecords while technically not publishing them, and therefore may be getting potential protection while avoiding statutory limitations. Professor Benjamin Kaplan has an article about this in the Pen Law Review in 1955 and since we’re at Berkeley, I’ll point out the Herman Selvin, the year before, a Berkeley Law grad in California Law Review, had an article about whether performance should count as publication in 1965 the Register offered a stark example. He says, today, a work could be thrown into the public domain or subjected to the limitations of the federal copyright statute if one copy of it is sold, whereas it could continue to be protected under common law copyright perpetually and without limitations if it is performed for 50 million people on television. I think we have to go up. We skipped that slide, so let’s move ahead to the next one. Great. The study presented three alternatives to deal with these concerns. First, just allow all classes of unpublished works to secure federal registration, federal protection through voluntary registration, which would not be a radical change, but would end the discrimination. Second, continue divided federal and state protection, but make the dividing line not publication, but public dissemination that is communicating a work to the public, visually or acoustically, by any method, in any form, permanently fixed or not. So that state common law would protect only works that had not been made public, perpetual common law rights would then only apply to private works that had never been disclosed. Alternative C was to eliminate state common law protection for unpublished works and provide federal statutory protection for all unpublished and published works from their creation. Now, of course, we know where this story is headed, but is perhaps important to see that the particular ending we got to was not inevitable. The commentators on the study expressed various views. Interestingly, the Copyright Office submitted three questions to Judge Learned Hand for his views on this study. As far as I can tell, it’s the only time they did that for any of the revision studies. And Judge Hand said he favored treating public dissemination as ending common law copyright protection and affording only limited time protection under the statute. On the whole, I am disposed to leave to the state courts the protection of undisseminated works, provided that some period be fixed after which they come into the public domain. He suggested 100 years after the work was created, or 50 years after the author’s death. The next step was the 1961 registers report, which we heard about this morning, which essentially recommended combining alternative a and alternative B, moving the dividing line between state and federal to public dissemination, but extending the privilege of securing statutory copyright by registration to all types of works. The biggest concern seemed to be the absence of statutory limitations on common law protection of unpublished works that were publicly performed, especially by being broadcast on the burgeoning television and radio medium or distributed in phonograph records. Particularly this was a concern about perpetual protection because of the lack of a duration limit, but also the fact that common law works were allegedly not subject to fair use, although, as far as I can tell, no case actually holds that fair use doesn’t apply to unpublished works. There are lots of citations back and forth to secondary authorities, but the for profit limit on the public performance right under the 1909 Act didn’t apply to common law works, and the compulsory mechanical license didn’t apply, which was one reason, at least we think that some musical copyright owners relied on not getting federal copyright so that they wouldn’t have to be subject of compulsory mechanical license. The initial decision not to recommend Alternative C seems to have been driven by concerns over private works. The report says we believe there are overbalancing reasons to preserve the common law protection of undisclosed works. We believe a fixed time limit for the protection of private papers against unauthorized disclosure would be undesirable. The right of privacy should be paramount in that situation. To put it mildly, the register’s recommendations did not go over well, so I’m not sure Ralph Brown was right that there was the least controversy about this by 1965 Kamenstein described the start and the length of the copyright term as the focal point of the opposition to the 1961 report. There was very little support for the recommendations in particular, very strong objections against the dividing line being public dissemination, which was criticized as nebulous, confusing, dangerous, amorphous, et cetera. And the overwhelming sentiment was definitely in favor of a single federal copyright system for all works, published or unpublished, disseminated or undisseminated. So this led to one of the shifts from the 1961 report to the 1963 preliminary draft that Peter didn’t mention this morning in 1962 Kamenstein announced that the copyright office was prepared to change its position on some debatable positions. In particular, the office contemplated revising its recommendations based on the concerns, and said, We came to the conclusion that the criticisms were generally well founded, and that the advantages in a single federal system outweigh the advantages of preserving common law copyright for undisseminated works. And unlike the 1961 proposal, the office proposed that existing works under common law protection on the effective date of the new Act would be brought into federal statutory protection and guaranteed a minimum term. Notably, the register pronounced that the office was now satisfied that the limited term of federal protection was sufficient for undisseminated works and private papers quote, we believe the terms provided are long enough to avoid any questions of invasion of privacy. So quite an about face in about a year’s time. And this was the approach of the office, starting from the 1963 preliminary draft and carrying forward. The principal advantages it identified to a unitary system where it would quote, clear up the confusion, uncertainty and capriciousness of current law and publication. I leave it to you to decide whether it did that, whereas moving to public dissemination would quote, make the current imbroglio of publication considerably worse unquote. It would further the basic constitutional purpose of national uniformity, and it would carry forward the limited times provision by getting rid of perpetual state protection. Not surprisingly, given the register’s description of this issue as the most fundamental one, the change had ramifications throughout the rest of the proposed legislation, which had to be addressed. They had to determine what if any national origin eligibility requirement would apply to unpublished works. There were some suggestions by some in the meetings that the same national eligibility origin requirements that were going to apply to published works should also apply to unpublished works. But the office. Listed that and carried forward the common law approach of protecting these works without regard to nationality. Requirements were adjusted. Adjustments were required for section one oh, what became section 105 which originally just carried forward the prohibition on copyright in US government publications. But now, of course the government would have unpublished works. And so 105 was adjusted to extend to unpublished works of the US government. Perhaps most significantly, the move to a unitary system affected duration. So first of all, it added to the Push to move to a life plus term. Jane articulated this morning, the comparative law reasons for this, but it was also buttressed by the fact that publication or public dissemination could no longer serve as a universal starting point for the running of a fixed term of copyright. But secondly, it required a section that is now section 303 substituting federal copyright for state copyright on existing unpublished works. Notably here, the section 303 originally brought into federal protection all existing and state protected works, including sound recordings. And this apparently was not worth mentioning in the discussion of section 303 although when they got to talking about the notice requirement, the register observed, it is not now clear whether sound recordings that have been published by public distribution of phonorecords are in the public domain or are still under common law copyright protection. The bill makes no effort to decide that question with respect to sound recordings published before the effective date of the new law, but takes account of the possibility that some of those sound recordings may be entitled to statutory protection after that date under Section 303. Unsurprisingly, the record industry was not happy about this, and in 1967 hearings testified against it and got people to make proposals on both the floor of the House and the Senate to amend the draft bills to make clear that published, previously published phonorecords of sound recordings did not get rid of the state common law copyright, and they would come in under Section 303 those were all unsuccessful, and it’s not until late 1975 that section 301, gets amended to provide that the 1976 act will not bring pre 72 sound recordings into federal protection, but leave them to state law. So the differential treatment of pre 1972 sound recording seems to be a consequence of the unsettled state of common law at the time that we decided to tear down the wall and bring unpublished works into federal protection. Okay, after the shift in the 1963 preliminary draft this basically remained the foundation of the revision bills, and, of course, became the law on January 1, 1978 so we want to think briefly at least about the consequences of the move to federal protection for unpublished works. So in some ways, the substantive law governing unpublished works didn’t change at all in 1978 copyright protection arises automatically upon creation. It does so without the need to comply with any formalities, and it does so for authors of any nationality. It’s just that it’s federal copyright that arises, rather than state copyright. And publishing a work still required placing a proper copyright notice on every published copy, as Tyler pointed out, the elimination of that doesn’t come until 1989 and required mandatory deposit in the Copyright Office for the Library of Congress. Other aspects of the substantive law changed dramatically, perhaps most significantly, duration, instead of being protected indefinitely and then for up to 56 years, when published under the pre 1976 act system, we get the limited duration of life plus 50 or creation plus 100 this creates what I have previously called the unpublished public domain. For the first time in US history, unpublished works go into the public domain, starting in 2003 which I think was probably the largest single deposit ever into the public domain. Every unpublished work of every person who had died before 1933 suddenly became public domain. And of course, now it’s every person who died before 1955. With respect to transfers of copyright, they are now divisible even for unpublished works and they require a written instrument to transfer. Those transfers are subject to termination under Section 203 and they get the benefits of recreation under 205. None of that was true of the state common law system. All of the statutory limitations apply in sections 111 to what was 118 when the 1976 Act came into focus. So government works are unprotectible under 105 even if they’re unpublished, fair use applies unequivocally, even though it was unclear whether it didn’t apply before. Library and Archive copying in 108 is allowed. Performance and display limits in Section 110 apply and the sound recording performance limits and sound alike limits in Section 114 all apply even to unpublished sound recordings. Enforcement changes because, of course, now registration is a prerequisite to suit for infringement of an unpublished work, if you’re a US work which was not required under state law, but timely registration also provides prima facie evidence of validity and ownership. And of course, we now have exclusive federal court jurisdiction over infringement claims, as opposed to state courts in the absence of diversity jurisdiction, and the remedies are better in some ways, injunctive relief is now federal and therefore more easily enforceable nationwide, impounding and destruction is available, and statutory damages and attorneys fees are available after registration. But perhaps most importantly, having federal copyright in unpublished works mostly made clearer and nationally uniform the law that applies to and governs those works, answering many of the questions about copyright, common law protection that were previously unclear and uncertain in light of the relative paucity of case law and the fact that there were multiple jurisdictions, so the law might be different from state to state. Now, everywhere in the US, unpublished works are governed by Title 17 and its interpreting case law, which has clarified the uniform definition of publication, the originality standard in 102 aAapplies with full force. The idea fact system distinction 102 B applies with full force despite some suggestions, for example, in the Prince Albert versus Strange case, that merely describing unpublished etchings by the Queen would violate the common law copyright, the rights of the copyright owner are only those granted in Section 106 again, despite suggestions in the case law that any use of a work protected by common law copyright would infringe the transfer of the material object in which the unpublished work is embodied does not itself transfer copyright, despite many jurisdictions having cases which said the unconditional sale of the original carried with it the common law copyright. Infringement requires a showing a factual copying and improper appropriation, unsatisfying as that test might be, as Oren has explained, unclear entirely before that, what exactly the test for infringement was under any particular state law doctrine, doctrine such as Scènes à faire and merger apply to unpublished works, and the joint work and work made for higher status of unpublic works, of unpublished works, is governed by the same uniform tests as for published works. So a pretty substantial consequence of bringing all of those unpublished works into federal protection. I want to close by just looking at some subsequent developments. So with respect to registration, it turns out that there’s continues to be a large amount of registration of unpublished works. The annual reports of the examining division no longer provide the breakdowns after 1977 and the Copyright Office uses different categories under the 1976 act. But basically 38% of all registrations from 1978 to 2022 are for unpublished works. Perhaps most significantly is the first line there where monographs in the non dramatic literary works category which were not registerable as unpublished works under the 1909 Act 27% of all registrations for monographs are for unpublished works. So clearly the copyright office was right that there was a demand for registration for that category of work. And of course, these registrations of unpublished works are no longer necessary to secure federal copyright protection, as they were under the 1909 Act. They just get the authors or copyright owners the additional advantages of registration. So it’s interesting and perhaps telling that we still have such a dramatic proportion of registrations for unpublished works. Finally, it turns out that the work of the legislature was not done with respect to unpublished works when it passed the 1976 Copyright Act. Four of the nearly 80 subsequent amendments make adjustments to this feature of the 1976 Act. So of course, we amended section 107 in 1992 to make it clear that fair use applies to unpublished works after the Supreme Court decision in Harper and Roe made it unclear exactly how that was to work out. Again, the Harper and Row court said, quote, fair use, traditionally was not recognized as a defense to charges of copying from an author’s as yet unpublished works, but it only cites secondary sources, none of which, as far as I can tell, actually cite any holding of any case. But now it is clear that unpublished works are subject to fair use. We amended section 303, to clarify, this will sound familiar to you if you’ve been listening, that pre 1978 sale of phonorecords was not publication of the work recorded on the phonorecord. Right, because after the Second Circuit decided that issue in 1976 the Ninth Circuit in La Cienega versus ZZ Top in 1995 held the opposite, that sale of records did constitute publication of the work recorded. I have a fondness for this decision because I may or may not be married to someone who may or may not have been a clerk for the judge who decided the case. But basically, the Ninth Circuit took the position that the courts in the 1950s in dicta had indicated that had caused such an uproar in the copyright bar. And so Congress overturned that position in Section 303 B, first in 1997 as to musical works, and then in 1910 sorry, in 2010 to literary and dramatic works. And then finally, we brought 1970 pre 1972 sound recordings into ui generis federal protection under Section 1401, in 2018 bringing us full circle back to one of the concerns that motivated the move to a unitary system in the first place, and I would argue, representing the final removal of the last bit of rubble from the 1976 acts destruction of the wall between federal and state copyright protection based on publication which had left those recordings to state common law copyright. Now I think there is nothing of fixed, unpublished works left for state law to protect by common law, copyright, though, maybe by other kinds of means of state law. Thank you.

Marketa Trimble  32:43

Good afternoon. Thank you for the invitation to speak about US state statutes on copyright. Of all things I have ever written that I thought would be useless, this was probably number two on the list, so I’m delighted somebody still cares. So welcome to the remarkable, very underestimated and sometimes somewhat bizarre world of state law, state statutes, and particularly state statutes related to copyright. As actually Christopher happened to mention earlier, there are, of course, plenty of state laws that affect copyright and copyright law, be it law on employment, labor, of course, contract law, inheritance law, divorce law, plenty of state laws affect copyright law, but today I will focus on state statutes that are specific to copyright in some sense. And as I go through these various examples, some of their, some of them are more entertaining than others, do keep in mind what one might expect of state laws in general. So we would probably expect that they might reflect some specificities of individual states, some diversity, some diverse priorities, challenges, preferences. We might also hope that state laws would provide this wonderful legislative laboratory in which native solutions enrich the scope and variety of possible approaches and something for the future, maybe for Congress to take on. Now, to what extent this is indeed the case is, of course doubtful, as we have stakeholders with nationwide interests, and also we might doubt this narrative about native solutions, given that we of course see how legislative ideas spread from one state to another. So given what we just heard, we might wonder what’s really left to state statutes on copyright, if anything. We know that even though there is no field preemption in copyright, many state statutes specific to copyright have been held preempted under Section 301, often it took long time for them to get held preempted, like the California Resale Royalty Act from 1976 it took until 2018 for that act to be held preempted by the 1976 Copyright Act. Of course, section 301, helpfully tells us, at least that unfixed performances can be governed by state law, and that state statutes will not be preempted if they concern unfixed performances. But also, then you have the government edicts doctrine and the recent decision in Georgia versus Public Resource that made clear that not only judicial decisions, but also legislative materials might not be covered by copyright. As judges and legislators may not be considered the authors of the works they produce in the course of their official duties as judges and legislators, regardless of whether a given material carries the force of law. So what is really left to state legislators to do in terms of copyright legislation? Well, first of all, if you look at state statutes, you realize how outdated most of them are. They typically, in terms of what state owns or what state works carry copyright protection, use this outdated language about copyrighting, causing to be copyrighted, acquiring copyright, securing copyright, really only Colorado caught on that we no longer require registration for copyright protection. We still have plenty of laws out there, state laws that talk about the state agencies being in charge of copywriting various legislative materials that are clearly not to be protected by copyright. We also have state statutes that talk about copyrighting other works, for example, works by the county board of education in California, data processing software created by government agencies. And in Pennsylvania, for example, a specific statute talks about copyrighting geological and topological surveys of Pennsylvania. And then there are some really cute examples, such as the Arkansas history book that should be copyrighted official insignia for farm products of Maryland, whatever that might be, highway maps of Ohio and Oregon, Blue Book, which is really a phone book, so really interesting examples of what state legislators think is worth copywriting and that there should be an act of copywriting. Now, not all states are completely behind. In terms of the Georgia versus public resource case, there are some states that actually prohibit the act of copywriting. Some of these legislative materials that, again, under Georgia versus public resource should not be protected by copyright. Interestingly, in California, the state would also not claim copyright in consumer information publications relating to earthquakes that might be helpful. And certainly since Georgia versus Public Resource, a number of states introduced legislation to change their state laws and make clear that again, their state agencies will not claim copyright in legislative materials. So that’s really encouraging. An interesting example is Virginia that actually has in its statute provision under which materials that are owned by the state of Virginia can be licensed under Creative Commons or Open Source Initiative, so they specifically refer to those vehicles to make these materials more widely available, even if they are protected by copyright, whatever copywriting means in Virginia. So what other functions, though, can state laws have in terms of copyright? Okay, clarifying which state produced works might or not be protected by copyright that might be useful. There is actually a function for state laws to support the federal Copyright Act. Section 1401 specifically refers to state law as the law that should govern the rights owner, the rights owner to the rights to pre 1972 sound recordings. So it is at the very end of section 1401 so if you have never gotten there, do it’s section 1401 L paragraph 2, under which the rights owner, for the purposes of this new right, or not so old right, is the person that has the exclusive right to reproduce a sound recording under the laws of any state as of the day before the date of an enactment of section 1401. The any state is really remarkable, because the entire reason for not having the statute, the federal act specifying who the owner of the right will be is the recognition that state laws vary in who the owner of this right to reproduce a sound recording should be. So if the state laws vary, one would hope that we would now have some kind of choice of law provision in the federal act to make sure that we don’t have multiple potential owners of section 1401, rates. But we don’t. We don’t. So there are differences among state laws. Of course, in most of the cases, it will be the label, whether it is framed that way in state court decisions like in California or specifically in some states, the state law talks about the owner of the master disk, so that should clarify it. But there are some states like West Virginia whose laws is different, and therefore we might end up in an interesting situation where we could have multiple, apparently, owners of rights under Section 1401, with no choice of law provision in the federal act. So that was maybe on the less humorous side of state law. So what else, state law can also enhance the protection of authors, believe it or not. So we already heard earlier today about the new, relatively new California statute on AI training on copyrighted materials and the transparency provisions in that act. But there are some other examples of recent bills presented in Maryland, for example, that have interesting provisions that suggest that state law could protect authors. So for example, there is a bill in Maryland presented right now on commission public art, artist rights and registration of art, which is a bill that would prohibit the inclusion in a contract for state public art of a requirement that an artist waive moral rights under VARA, which you know could be in favor of the author. It could also not be depending but the act, the bill, does have provision in addition to that, that the artist may still choose to waive the VARA rights if the artist wants to. Another remarkable bill in Maryland is a bill protecting artists creative expression, the so called Pace Act, which would impose a limitation on admissibility in criminal proceedings and juvenile proceedings, of uses of creative expression of a defendant or respondent. So it would protect artists against their own works being used against them in criminal proceedings. It’s rather a remarkable bill that suggests that the court would have to figure out whether the defendant or respondent intended the creative expression to be literal rather than figurative or fictional. Also, it suggests that the works, while they could not be really they wouldn’t be admissible in evidence for mens rea. They could be used to decide on evaluating, recommending or ordering referral to mental health services or diversion programs. So clearly, something that came up in the context of particularly rap music, and the question whether, when the artist says something in that song, whether they really mean it. So that’s a bill that’s still pending in Maryland, but the state laws could also protect users. So not only authors, also users. There are acts that, of course, clarify the question of access to public records and the misconception that copyright could be somehow used to prevent access to public records. Lately, we have a series of new state bills on the topic that talks specifically about requiring. Public access to learning materials, which is, of course, related to the recent wave of opposition to certain things being taught in our schools. So that’s part, part of those types of bills and acts. Many of you might be familiar with the state acts concerning digital lending of E books, where the original bill in Maryland was adopted but then struck down in courts, and now Connecticut became the first state to adopt since then, a bill that seems viable. It protects, again, digital e lending and prohibits libraries in the state of Connecticut from entering into contracts or license agreements with publishers of electronic books and digital audio books that contain certain restrictions. So while the Maryland bill really limited contractual freedom of the copyright owner. In this case, this bill in Connecticut focuses on the libraries and their freedom to enter into some of these contracts. Two more examples, demand letters, so similar to some of the demand letter legislation in patent law, some states have considered bills that would spell out what a demand letter should look like, so that no abuse of rights would arise. And my particular favor is the last one, because it’s an act that was adopted based on a bill presented by one of my former students. I have nothing to do with the bill or the act. It is an act in Nevada creating a requirement for law enforcement agency to adopt written policies and procedures governing performance of copyrighted works by Peace Officers while on duty. I have to say that the agency in charge of training of Nevada officers was puzzled in its statement, it said the language of the bill is so broad that the agency cannot determine the intent of the legislation and specifically in what circumstances a peace officer might be performing copyrighted materials. You might recall it was that story about the LA Police playing songs really loud out of police vehicles so that the bystanders who were taping them in action could not post those recordings on YouTube. Okay, that bill takes care of it. So that’s that’s law in our state. Okay, so that was state laws protecting users. What else could state law do? Of course, state laws could be really, truly innovators in law. Now, to what extent this might cross the line and become a subject to preemption is an interesting question. So, of course, in Arkansas, we now have an act on generative artificial intelligence tool, specifically on ownership of model training and generated content that talks both about ownership of generated content by the person who provides the input or directive to the generative AI tool, and also about ownership of the trained model by the person who provides data or input to train a generative AI model. In New York, there is a bill that’s somewhat similar and yet different from the one in California in terms of transparency. It’s really more geared towards protection of journalism and journalism providers, and it’s about disclosure obligations as to crawlers. So not only training data for AI generative AI models, but also for the crawlers. My last example of what state laws can do for copyright is fascinating. State law can enforce your copyright outside of the United States, and that’s really a fun example. Now we don’t know how these state statutes would would do after the RJR Nabisco decision by the US Supreme Court and all that. But, you know, it’s an interesting example. One would think that copyright is really territorially limited. Of course, we have the wonderful, underused vehicle of the predicate act doctrine, or constructive trust theory, under which in some circumstances, copyright owner can collect profits that arise from Acts outside the United States as long as there is a predicate act of infringement in the US. But the statutes here on the screen are interesting because there are state statutes on unfair competition in. That suggest that it might be possible to sue in the United States for an act of unfair competition that’s based on an act outside of the United States of some kind of intellectual property right infringement, and indeed, in few examples in which state agencies turn to these laws, they did apply US law to the acts outside the United States. So if somebody gains advantage, let’s say in the US market, by undercutting prices for solar panels, you might look at that company’s use of IP in outside the United States, let’s say, China, and then apply US copyright law to those acts in China to find out whether it happens to infringe US law in places where it shouldn’t apply, and that would be then held to be unfair competition law under state law. So maybe the most remarkable example of what state law might be able to do for you. So again, one has to wonder whether we will see more interest of state legislators in legislating on something related to state law, we might wonder whether there is some increase in state legislatures activity on this topic. Of course, it could be related to the problem we have and more than one with Congress. But it also can be that state legislators are becoming more ambitious. Maybe another reason is that there is a greater interest among the public in copyright law now that we are all users and also authors at the same time, perhaps it is more appealing for state legislators to try to enter into this largely preempted world. Thank you.

Guy Rub  51:59

All right, so I’ll talk mainly about the relationship between breach of contract claim and federal law and copyright law. I want to start by thanking Berkeley and the organizer in Columbia for having a wonderful event and for including me in it. So we know generally there are multiple type of preemption. There are mainly four main ones, express preemption and field preemption, actual conflict and obstacle conflict, but two of them are not really relevant to our topic. So I will talk about express preemption and about obstacle conflict. Today, we start with this quote from the House report that called Section 301 one of the bedrock provision in the act. Although if, if you read the rest of it is more relates to what Tony was talking about when they are talking about the art of section 301. So section 301 is a broad provision, but much of the discussion of it in the legislative history revolve around a common law copyright and what to do with that. In the legislative history, there was another provision there, subsection B, that included a list of causes of action that won’t be preempted, included breach of contract, and also include the one sentence in this entirely state of history that talks about the topic which has nothing the bill derogates from the rights of party to contract with each other and to sue for breach of contract. The difficult part is that provision was not included in the final bill. I know Shaym will talk about that. So a so what do you do with that is slightly  controversial and court was struggling with, what does the meaning of removing that provision? So we are left with section 301 that is mainly 301A, two requirements. One is that the has to be the subject matter requirement that the work that the contract we are talking about, whatever the cause of action may be, has to do with subject reading, the subject matter of copyright. And it is well established that that requirement should be read very broadly for interest of time. I’m not going to go deep into that, but the second requirement is the one I’m going to talk about, and that is the equivalency requirement the ACT expressly preempt only causes of action are equivalent to the exclusive right under Section 106, and if you look at the legislative history, there is this wonderful quote about, we thought the house thought that this will be a very, very easy and very very trivial. You want to make it in the clearest and most unequivocal language possible to foreclose any conceivable, conceivable misrepresentation, to avoid the development of any vague borderline area between state and federal protection. If I have to reflect, to summarize what I’m going to talk about with exhumidities, oh boy. It is extremely, extremely complicated, and the idea of not creating a gray line, great. It’s fantastic if you can do it, but we are not even close to not creating a gray line. Most of what we are talking about, I’m going to talk about for next few minutes, is going to be a gray line. In other words, if the federal, if Congress thought that, you know what can be so unclear about equivalent. Equivalent, if it’s equivalent, is preempted. It’s not equivalent. It’s not I would argue that it’s actually very, very difficult. It’s difficult because there is a, Marketa also mentioned that in the previous panel, Chris talked about that there is sort of an entanglement between copyright and state law, and especially state commercial law. We have a system that seems maybe, if you look at the copyright on its face, that seems to be holistic, but it’s not. It can operate. It can it get great, exclusive rights, but those rights need to be transacted. You need to have commercial law to be able to sell and buy those right. And you can see it with the also definition, we talked about in previous panel, how the Copyright Act doesn’t define some stuff. Think about all the terms that Copyright Act used from commercial law without defining, like sale or assignment or mortgage or my favorite, license. Those are all term of art in commercial law that the Copyright Act used quite casually, repeatedly, dozens of time without bothering defining but this is not like no defining authorship, assuming that that definition will be fine found in copyright case law, I think we left that open in the assumption that state law, basically commercial law, will fill that gap. So we already have a very tangled system, but that entanglement also creates some conflict, because while copyright law regulates what right and a limitation we have on a fixed creative work, contract law can create the same thing, they can also create a norm about what we can and cannot do with the same work. In fact, it could go much further than that, in some respect, and I know many of us think that copyright law is too broad, but we have limiting built in mechanisms that are so, I think, banal, that we don’t even think about that. But copyright law cover only certain subject matter and not others. It covers certain action and not others, and it’s subject to all the defenses in the ACT contract. Contract law is not subject to any of that. Contract law can absolutely define right in the same subject matter or in additional subject matter about the same action on additional action, and it can ignore the defenses, and it sort of relate, I think, the question of how much should we do it related to also this question of how do we think about copyright law itself? Here is to quote un Supreme Court, and there are dozens of quote like that, bottom Supreme Court and other cases. So there are line of cases in which we think about copyright, and we talked about copyright has created this delicate balance between competing interests on the on the public interest. And I think if you think about copyright this way, you will be more bothered with state law that sort of mingle with that, intervene with that, and changes that delicate balance. But we also have a long line of cases that define copyright as a property right. And if we think of copyright law as that, I think we will be more open to the idea that if people have copyright property right, they should trade on their property right. So what’s the issue? So I’m going to talk about the old status quo, and then I’m going to talk about what I actually asked to talk about, which are the new development that we have in this area on the law in the last few years, and the new frontier that those cases are being litigated in the last few years. So the all status quo, I want to talk about the decision that was were decided up until a few years ago. First, a caveat, I’m talking about, of course, decided opinion, and we always need to remember that for every decided opinion, there are many, many disputes that are not being decided. So whatever I’m telling you now, it’s, by definition, a partial reflection of what we have seen in the case law. So there are about 350 cases in which, 350 opinions, in which courts decided whether a contract is preempted by section 301, of the Copyright Act. Practically all those cases are between businesses or between professional I don’t want to it’s not business in the technical term. Sometimes it’s repeated player within the copyright industry. But those are people, those are not the type of user that you know click I agree, without us having any expectation of what they are doing, I’m not saying that that doesn’t happen. That’s not what you see in the in the litigated cases. By far, the most common fact pattern that we see there is ID submission by a lot, about, about 25, I don’t have the exact number, but about 25% of the cases are ID submission. So someone submit an ID, so no protected by copyright, and the question is, can we limit that submission with a contract? And some of those contracts are implied and some of them are expressed. Those of you know it’s called Destiny claim. We call them, many of those claims, especially in California. So this is by far the most common one. The uncommon one is all over the border. We see, once you go beyond any submission, we see few cases in every category. We have ton of cases about database protection, basically limiting the distribution of databases. We see cases in which we basically have a longer statute statue of limitation compared to the Copyright Act. So we see a real variation of cases. Cases around fair use are actually extremely rare. They got a lot of attention because there are a few cases in which we we have the contract that limit reverse engineering. And those two cases, there are two cases just went to an appellate court, and they got a lot of attention. But other than those two there’s actually very few cases, there are very few cases in which the contract is a contract around around fair use. So moving to talk about the law, what courts use routinely, both within contract law and outside contract law, to decide whether a state cause of action is equivalent to a copyright law is what they call the extra element test, and it’s a rather formalistic test, I would say, at its core though, the element of it that allows some flexibility. The idea there is we know what the element of copyright cause of action, we can list the element of state cause of action, and we see if the state adds something additional. This is the extremely formulaic side. The addition to it courts also say not every technical difference will suffice. So the difference is supposed to create actual difference, actual meaningful difference between the state cause of action and the Federal cause of action, they might need to change the nature we are saying. The nature of that extra element need to be has to do with the nature of the actual cause of action under a state law. In the case, when we see two, version two, sorry, two approaches, and both of those go to the question whether a promise is that extra element. Because I told you what we are looking for, an extra element. So question, can the promise be the extra element? And a minority of court says no, at least not categorically. In other words, if the promise is a promise to refrain from what of the action that is listed in Section 106 so if it’s a promise not to reproduce or not to adapt or not to distribute or public performance or public display, then it is preempted. So that’s the minority approach. The other approach is that the promise is the extra element. So if the promise sometimes they say the consideration is the extra element for our purposes, It leads to the same place, because every contract. When I talk contract, here is a side note, I’m talking actual contract law, actual cause of action in contract. So not quasi contract or no implied by law contract. We can talk if you want, go down that rabbit hole, but those are not contractual cause of action. But if you have a contractual cause of action, there is always a promise, and there is always a promise and there is always consideration. So if that your a guiding principle to how to find whether there is an extra element, the result will be there is an extra element, and therefore the contract will not be preempted. Now that approach is only mostly identified with proceeding. We already mentioned, It’s already mentioned in one of the panels this morning It actually did not start with proCD. It’s not even in appellate court start with proCD, but proCD is by far the most famous case in which that approach was listed. It’s heavily criticized in the literature, although when you look at case law, it’s extremely popular, and most Appellate Court adopted that one way or another, with minor variation, and until recently, the only Sixth Circuit is the only circuit court that says we explicitly reject proCD. There are new developments here. We’ll get to that in a second. Why did proCD win the day? I’ll try to be brief here. There’s a lot to say about that. I can list two reasons. One, it’s easy to apply, It’s a bright line rule. And the second one is that, and especially when you read, I read all those opinions, If you read those opinions, especially those opinion that adopt that minority approach, what I call, or I call it in some of my work, fact specific approach.  You see, they are extremely inconsistent. They are messy, and they are partly messy because I think the test itself that they are using to find what is accurate and what is not is both under exclusive and over exclusive. It’s under exclusive because there are many contracts that we will find troubling, I think that are technically not about recollection of reproduction. Every contract, there’s a payment. There is a line of cases, a long line of cases from court that adopt the minority approach that says, as long as the obligation is obligation to pay, go, go, do your thing, because payment is not mentioned in Section 106, now that’s a huge loophole around a lot of thing you can do with payment. There is a line of cases in there says, the broader the contract, the more easy it is to escape preemption. So if I limit reproduction, no, no, no. But if I limit reproduction, plus another thing that is not section 106, it’s a void preemption, which is counter intuitive, it’s also over exclusive, because we have a lot, a lot of contract that we think are non problematic that limit reproduction and remit a public public distribution, public display of copyrighted and non copyrighted work. You think about a non disclosure agreement. You think about the standard way in which we protect databases, (inaudible). So the test, I don’t think the minority approach came with the test that is great divider. I won’t talk about it much. Also, in some of my work, I suggested that I don’t think I think those who argued contract will just replace copyright. I think it’s oversimplification. I think that built in limitation in contract law itself, like the limitation of remedy and like no secondary liability. From the good old days we had secondary liability on the other side. There are limit, built in limitation in contract law itself, that doesn’t allow it, I think, to replace copyright, regardless of preemption. Why? Let’s talk about modern development. And there are two of them, one on each coast. So the first one is the Second Circuit, and both of them, interestingly, deal with pretty similar fact patterns. They are both dealing with data mining. They are both dealing with the situation that the website has information on their web, on the website probably not protected by copyright, or they are not the copyright owner. And the question is, can I scrap it? Can I data mine it and copy it, or not when the term of use says no? So one of them deals with a genius, it was a website, still a website that hold lyrics to a song, but if you are trying to look for lyrics on Google, you know, you write the lyrics, the lyrics are on Google. You don’t need to go. So genius, I know it’s hard to say, but genius is right there at the bottom. How did, how did Google get this in those lyrics, in order to put it on Google and push genius down, which completely destroyed the traffic to genius, which is their business model. They got it from genius, they scrap genius website to get those lyrics that pushes genius, basically drives genius almost out of business. So what can genius do? They can’t use copyright because both of those companies are licensees. Genius has a license and Google has a license, so they are both licensees. So they cannot use copyright, but they have a browser app, and the browser app says that you cannot do that. You cannot use our website in that way. They sue Google for breach of contract, and Google argues preemption, and they won. They won in the Second Circuit. The Second Circuit said, and it’s interesting enough, because the Second Circuit, for decades, somehow was able to avoid the decision on this matter. Although we had about 50 to 60 decisions from, just from the district court in New York, the Second Circuit in sometime intentionally said, I don’t want to decide this case. But then in an opinion that almost did inside the other opinion did inside the circuit split in a very, very short opinion, the court says, well, you’re just trying to regulate the public, public display of copyrighted material. You can do that in a contract. This is preempted. Genius, of course, tried to erase the proCD line of cases, but it didn’t work. Interesting enough, the case runs the Supreme Court, and the Supreme Court asked for the opinion of Solicitor General. They rejected the search eventually, but the Solicitor General made a completely different view, which says the issue is the browser app. Because it’s a browser app, this is not a good vehicle for the Supreme Court, pretty clearly implying that in the although they didn’t say that, clearly that in the opinion of the Solicitor General, browser apps are preempted, while anything else should deserve different treatment. Now this is, of course, it’s really important, because first of all, there, we have a ton of websites that has non protected information on them that include factual information. This is supposed to remind them, there was a discussion in Congress about how search engines are basically killing those websites by taking their information and presenting them when users search for it without actually drive traffic to them. We know that there are discussion in various states, including this one, and in other countries about what you do with news agency. Those are all the same kind of discussion and the same topic. And of course, the issue of data mining that has to do, what with copyright issues, copyright protected information, does it is not, is decided by that. And many of those, some of those cases, not all of them, were decided copyright cases. So LinkedIn has as a very famous case against hiQ labs and hiQ lab won on a lot, but at the end of the day, LinkedIn haa contract, and they just won as a contract cause of action. The last development I want to talk about is the alternative, and this is to rely on implied preemption instead of express preemption. So I think the bottom line, if it wasn’t clear by now, I think section 301, asked the wrong question, asking if they are equivalent, slightly misses the mark. Contract can create a lot of arm to our copyright policy without trying to replace it equivalent sort of imply we want to channel what is copyright to the copyright system. That’s not the issue with contract. The issue with contract. They might be an obstacle to the to what we are trying to promote with copyright policy, and in that case, we might want to apply conflict preemption. There is an extremely, extremely thorough analysis by the Second Circuit, just a few years ago, by Judge Leval in a case called In re Jackson, it has to be the right of publicity. He uses a lot of academic work, including by Rebecca Tushnet, including by Jennifer Rothman and by others, to come up with a test. And he said, we need to look at two questions to know if there is this type of preemption. We look at what value the state is trying to promote. Are they different? So if the state tried to promote privacy, for example, so in this case, if this a contract that tried to promote privacy, then it’s probably different enough from copyright. If we try to promote commercialization of information, we have a bigger issue. And then you need to look about about whether there is actual conflict between those situation. There is one case, a really important case in my mind, from two years ago, X Corp v. Bright Data that applied this approach. This is, again, this is Twitter trying to protect a company from scrapping their services. And the court held that this is preempted. The really by conflict preemption, the really interesting situation there is that how Judge Alsup phrased the arm to the copyright policy that is intuit(inaudible) policy. He says there are two kinds of arms, or two main kind of arms. One is that you, Twitter, do not own the copyright for those posts. How can you limit anyone from commercializing their own copyright, which is true in theory, although I don’t think any Twitter user is trying to commercialize their post. The other one, I think, is the real kicker, he says, might have fair use claim here to that mining. So how can you have a contract that don’t allow others to use their fair use? It’s very close to using fair use as almost a right, and I think it’s foreshadowing a decision that you will come to a year later with AI data, with cloud. Finally, I would say, one of the reasons that I support conflict preemption, I think it’s a better framework. I think it’s allow us to ask questions that are more relevant than the formalistic is it is the equivalent or not? We can ask the question, was it in negotiated contract or not? Which is the other approach, almost ignore. We can ask ourselves, is it the content to try to promote privacy or other interests or commercialization? We can ask ourselves, what market power do the parties have? I think, all those questions and many others, is it is it a researcher that’s using the the work? Or is it commercial use? You can open the door when you create a flexible framework like conflict preemption versus section 301, that is a sort of more strict. That’s that. It’s a difficult question. I think that the reply, the assumption we won’t have a gray area, was nice, nice idea. I think we will have a gray area by definition. I’m not sure that section 301, is the right vehicle to tackle that gray area. Thank you.

Shyamkrishna Balganesh  1:14:35

Okay, I’m going to continue forward on the conversation about preemption, but talk about an area that might seem absolutely minuscule in terms of its relevance. And let me start off by telling you why I think talking about the hot news misappropriation doctrine and its influence on the copyright preemption provision is important. First, because I think it had an outsized influence on the trajectory of the preemption provision. Itself, and I’ll talk about that, and the misleading nature in which the actual evolutionary story has been told and evolved comes from the story of misappropriation. And secondly, that misleading account continues to influence the manner in which courts develop the doctrine and actually talk about misappropriation to this day. So let’s start by talking about misappropriation. Here we go. Okay, so if you flash back to first year property law, everyone, hopefully has heard about the 1918 case of International News Service versus Associated Press, the genesis of the hot news misappropriation doctrine. To refresh your memory, this was a case of a conflict between two large news gathering cooperatives, the Associated Press and the International News Service. The allegation was that the International News Service was relying on the factual content from the news stories of the published Associated Press newspapers to actually recreate stories of its own, and the question was whether the court could invoke its equitable jurisdiction in order to provide a form of remedy. Now this triggers a majority opinion by Justice Pitney and a dissenting opinion, two dissenting opinions, one by Justice Brandeis and another, excuse me by Justice Holmes, but the majority opinion relies on unfair competition law and implicitly the doctrine of unjust enrichment, to develop this framework that comes to be called quasi property. And the logic for it is that free riding on the efforts of a direct competitor to appropriate information when that information has time sensitive economic value, such as news, can trigger the equitable jurisdiction of the court, right? So what happens is that the AP is able to obtain an injunction against the INS for such period that the news has time sensitive economic value, which is typically for 24 hours. The case produces a blistering dissent by Justice Brandeis. And in some ways, the Brandeis dissent is perhaps more popular than the majority opinion. And nevertheless, INS versus AP is a decision which developed the misappropriation doctrine, rooted in this term, quasi property. Now, what happens, though, in the next three or four decades by the 1960s the INS doctrine, which comes to be called the misappropriation doctrine, begins to see a drop in its influence. And there are two main reasons for this. One, the obvious civil procedure reason, which is Erie versus Tompkins comes to be decided, which says that federal general common law is dead, and effectively the jurisdictional basis in some ways, for the INS doctrine ceases to exist. But the second reason is more important for us, which is that a whole series of circuit court opinions, most prominently led by the Second Circuit and Judge Learned Hand come to criticize and limit the scope of INS versus AP. The argument being that it was very limited to its scope. The court had exceeded its judicial role, and in some ways, any intervention in the IP space needed to be done by Congress. So it was a question of institutional competence. But in reality, Judge Hand and and the Erie doctrine never really managed to kill the doctrine, because state jurisdictions, either statutorily or through common law opinions, began to absorb it. So it still survived, the scope remains somewhat unclear. So now let’s fast forward to the copyright statute. So if you begin with the House report and the Senate report, which obviously is the principal basis for thinking about the legislative history of the act, what you start seeing is that, very interestingly, misappropriation finds reference in the statutory provision around preemption contained in the House report, which is section 301B3, right, which is the only statutory provision that it represents over there. And this is from September 3, 1976, and so this is the statutory provision which talks about exemptions from the preemption provision, and it lists activities violating legal or equitable rights that are not equivalent to any of the exclusive rights within the general scope of copyright as specified by 106 and then lists specific ones that meet this set of criteria, including rights Against Misappropriation not equivalent to any such exclusive rights. To make matters worse, the House report contains this very important paragraph about why it is including misappropriation, a discussion of the delineated categories by specifically indicating that misappropriation is structurally very different from copyright does not, is not in any way synonymous with the working of copyrights exclusive rights, and in fact, would afford courts and state jurisdictions the ability to invoke equity, to go down new domains, which all seems totally fine if you just start with the legislative history. The problem, though, is that if you look at the actual enactment, the actual 1976 Act of the enacted date, October 19, 1976, this is somewhat well known, the actual provision 301 b3 has no such reference to misappropriation. So the House report was based on an earlier version. Everything after the 106 idea comes to be eliminated. This is a side by side comparison. What you have on the top is the version of the statute from the House report. Sorry, just going on and on. Here we go. So this is the this is the side by side version of the original version in the House report at the top, and at the bottom is the version as enacted under the October 1976 enactment. What you will see is that everything referencing misappropriation gets eliminated. And because I spoke without moving the slides, this is the portion that I mentioned in the House report. This long paragraph about the rationale for the inclusion of misappropriation, which talks about why it’s very different, the structural dissimilarity to copyright. But all of this is contingent on there having been misappropriation referenced in the statutory provision, which never makes its way into the final bill. So there’s a fundamental mismatch between the House report and its statutory version and the final bill. So what’s the story here, and what effect did this have? And this is what I’ll propose to tell you in the next five minutes. Okay, so the story begins with the 1965 bill, after the register of supplemental report, which takes the position that misappropriation is far too equivalent to copyright law. There is a fundamental fear that the misappropriation doctrine would swallow up the very working of preemption. So in this report, there is a statement to the effect that the misappropriation doctrine is the virtual equivalent of copyright. Now what you will see is that the version of 301B3 that was contained there represents all sorts of different common law actions, breaches of trust, invasion of privacy, defamation, but no reference to misappropriation, and that was a very, very conscious omission. But then what happens? What happens next is that the Department of Commerce intervenes in the 1975 hearing, and the Assistant Commissioner of Patents comes in saying that, look, the misappropriation doctrine, we believe plays a very important residual role in the working of the overall intellectual property system, because it allows courts to anticipate new areas for development and retain flexibility at the equitable level, specifically referencing what happened in INS versus AP, and wanting to preserve state flexibility in this direction. Okay, so this, the Department of Commerce introduces an amendment which produces what we see today in the House report, right, the version that comes to delineate misappropriation in that list and accompanying it as an explanation for it, the portion on the misappropriation doctrine and its structural nonequivalence that I talked about. But what’s very soon emerges is an intriguing position, which is it becomes apparent that the Department of Commerce had not consulted with other departments in the federal government. So there’s an interesting colloquy where there’s a discussion coming from different representatives who ask the commissioner, have you spoken to other departments? And I’ll just read out for you, Mr. Railsback representative, I find myself in somewhat of a dilemma, who actually speaks for the administration. There seems to be disagreements with virtually everybody. We have the Department of State disagreeing with everybody except on the manufacturing clause. And now we have the Department of Commerce that takes a different view. Does anyone purport to speak for the administration? And the Commissioner of Patents is only able to say our testimony only purports to speak for the Department of Commerce, right? So there’s a crack that has been planted early on, but then matters get far, far worse. What happens next? The DOJ, the Department of Justice, through the Assistant Attorney General for antitrust, is asked to weigh in by a staffer in the office of Senator Hugh Scott, from Pennsylvania, specifically on the misappropriation doctrine and the Department of Justice in the last minute, sends in a letter to the Senate and the House, but Primarily the Senate, directed to Senator Hugh Scott talking about how the misappropriation doctrine is the worst thing possible for the preemption provision, that it would effectively neutralize the logic behind preemption. It would effectively liberate states to do whatever they want and specifically misrepresent it, what the misappropriation doctrine actually stands for and how it works completely talking about how it was a doctrine that was directed at creating monopolies, how it has serious antitrust concerns that have not been explored, how it is far too expensive and is basically simulating property rights. And what happens is that Senator Scott makes a big stink on the floor of the House talking about how this needs a further amendation to Section 301, b3 and that this needs to be eliminated, which passes through. And then later on, interestingly, the house, without any further discussion, takes the Senate’s version at face value and incorporates it, which becomes phase three, which is the complete elimination of any discussion of misappropriation. So this is where the status of misappropriation is seemingly unclear, because what happens when there’s the modification from the Senate to the house, the house adopts the position that, oh, there’s really no change, as we have been promised by the other side that is taking the position that we need to eliminate this. The net effect is they bury the very, very logic for why the elimination sought to happen. And this is what is the result in 301 b3 which is the provision that we’re stuck with today. So why does any of this matter in the modern context, other than being an interesting story that different branches of government don’t agree. No new news there. Well, I think the answer is very, very importantly, first and foremost, that there is this fundamental mismatch between what is contained in the House report and the actual law in two respects. And this is very, very important one in the version of 301B3, that is in 941476, in that version that was not the version that was enacted. There’s a long list of different causes of action, including misappropriation that never was enacted into law. Secondly, the legislative history has an elaborate discussion of misappropriation, and its logic also has no bearing now in light of the fact that that was eliminated from the final version. Why this matters even more is because different jurisdictions, different courts of appeal that sought to resurrect the INS doctrine, all of which happens in the 1980s and the 1990s unfortunately, rely on the old house report that did not get updated after the elimination. Case in point, the 1997 Second Circuit decision in NBA versus Motorola, which explicitly says, let’s look at the legislative history. It’s so abundantly clear that they wanted misappropriation. Let’s take it, run with it, and resurrect it into a full blown doctrine. It only gets corrected in 2011 when the Second Circuit revisits it. But what happens in the intervening 14 years? It moves from New York to Ohio to Missouri to Illinois to California, all these jurisdictions now have case law that have indirectly relied on the wrong legislative history and have not been corrected. So where do we go from here? I don’t know, but where we do what we do know is that it’s wrong and it’s been placed or been relied on for quite some time now, and very importantly, the origins of this comes from a fundamental error in the House report and its belief that misappropriation was somehow wholeheartedly endorsed, when, in fact, the opposite was indeed true. All right, thank you very much.

Peter Menell  1:28:38

Well, I think we’re at the end of a very full day with a mic drop to really end. And I’ll say I thoroughly enjoyed the breadth and depth of coverage today. It It restores my faith in the very nerd like geekish work of academics who are trying to figure out these puzzles. It leads us to this embarrassing moment where we are aware of problems that have infected the law and not clear what to do with it. One of the doctrines that I find very troubling, which we haven’t talked about, but is kind of in the background, is the statutory stare decisis doctrine, which I call the shirking doctrine, which is that when the Supreme Court becomes aware of errors that have gotten into Its juris into its cases, it decides that Congress can fix them. And one of the things we learned is that Congress is very slow to fix these things, and so that’s the world we’re in. But I just feel that our role is trying to understand and decipher all of this is Important and will, I hope, lead to better, more resilient laws down the road. But I want to thank everyone, especially this panel, because coming at the end of a day like this is probably the hardest job of all. We have a great reception to follow, and we hope to see you back here tomorrow.