Tag: Copyright Office

USCO Issues 2nd Report on Artificial Intelligence: Copyrightability

copyrightability

“Where AI merely assists an author in the creative process, its use does not change the copyrightability of the output. At the other extreme, if content is entirely generated by AI, it cannot be protected by copyright.” – Copyright and Artificial Intelligence Part 2, Copyrightability, USCO –

Last week, the U.S. Copyright Office released Part 2 of a planned three-part report on copyright and adjacent IP matters concerning the use of artificial intelligence. The new report expresses the Register’s views about the copyrightability of works when they are produced in some way with the use of AI. In summary, the Office reaffirmed the doctrine that human authorship is required for copyright to attach to a work at all; that copyright should not protect expression created by generative AI; and that the use of assistive AI should not disqualify a work for copyright protection.

Before proceeding, it’s important to remember that the question of copyrightability, or “authorship,” with AI tools is separate from the legality of unlicensed use of creative works for the purpose of “training” these models in the first place. As argued in other posts, most machine learning (ML) with unlicensed protected works should be held to be mass copyright infringement and should not be exempted under the fair use doctrine. Nevertheless, on the assumption that AI tools for creative work will continue to exist, the question of copyrightable authorship with these technologies is an important and ever-evolving doctrine.

Generative AI (GAI) and Copyrightability

The most difficult copyright question regarding generative AI (GAI) concerns works made with a combination of human-authored and AI-generated expression. As the Office report emphasizes, the question itself defies bright-line guidance because it is inherently a case-by-case, fact-intensive consideration that can only be weighed in the courts. That said, the report expresses a general view that GAI apps do not presently allow the user sufficient control over the expressive results to claim ownership in the outputs.

While the Office recognizes that selection and arrangement of GAI material can meet the threshold for copyrightability, and it leaves open the possibility of technological advancements to enable greater “control” of GAI tools, the report argues that GAI is presently a “roll of the dice” as a creative process. “No matter how many times a prompt is revised and resubmitted, the final output reflects the user’s acceptance of the AI system’s interpretation, rather than authorship of the expression it contains,” the report states. Acceptance is described as “authorship by adoption,” which is roughly the equivalent of claiming copyright in a work one finds rather than creates.

What this means as a practical matter is that creators may claim protection of their expressive contributions to works that include GAI material, but the latter should be considered unprotected and, therefore, disclaimed in a registration application. We shall see whether the courts agree with the Office, most immediately in the case Allen v. Perlmutter where Jason Allen argues that the nature and variety of prompts he used for his visual work were not like rolling dice but were instead deliberate steps toward creating his mental conception of an image.

Regardless of how Allen is decided, it will only be the first major litigation addressing the mixed human/AI question at issue. This highly subjective consideration will remain a case-by-case matter for the foreseeable future, even if certain GAI apps provide greater “control” for users per the Office opinions.

Assistive AI (AAI) Does Not Limit Copyrightability

“The Office agrees that there is an important distinction between using AI as a tool to assist in the creation of works and using AI as a stand-in for human creativity.”

As a creator, I appreciate the Office distinguishing GAI from assistive AI (AAI) and stating that the latter should generally not disqualify works from copyrightability. For instance, if one uses AAI to expedite color correction in a group of photos or to more efficiently check and make grammar recommendations for a manuscript, one need not disclaim the use of AI in these contexts. Likewise, it is important that the Office recognizes that AAI used within a larger work (e.g., to fix a scene or create an effect in a motion picture) is not a basis to limit the protection of the whole work.

While there may be lines inevitably crossed (e.g., an AI suggests, and the writer copies, whole paragraphs in a text), this would arguably be a case when AAI becomes GAI. Nevertheless, resolving protection in this gray area of authorship is likely a matter best left to the courts and not a line easily drawn by the Copyright Office. In practice, even if I did use AAI in my own work, I would not disclaim that use in a registration application, but if I allowed AI to truly write some material, I would disclaim that and not submit a fraudulent application.

Creators should remember that under Unicolors v. H&M, an innocent error on a copyright registration application is not a basis to void the registration. It is important to make a good faith effort to claim the human-made expression and disclaim the AI-generated expression, but the Supreme Court set a precedent that creators should not be penalized for an imperfect understanding of difficult questions of law when submitting an application.

It is understandable, of course, that creators want certainty, but in this report, I think the Office provides sound guidance for the moment while cases like Allen work through the courts. It would not be acceptable to simply default to protecting all GAI material while so much “AI slop” floods the market and, among other things, threatens to undermine the incentive purpose of copyright. For the author using AI in conjunction with her own talents and expressive capacity, we are at the leading edge of this discussion. For context, publishing has existed for a few centuries, but defining “publication” in U.S. copyright law still defies bright-line definition to this day. Hang in there.

What Does the Valancourt Decision Mean for Most Creators?

As discussed in an earlier post, Valancourt Books, a small, on-demand publisher, filed suit against the Copyright Office and the Department of Justice and argued that the Office’s demand for physical copies of published books is unconstitutional. Valancourt alleged that the authority granted by §407 of the Copyright Act to demand the copies (or be fined) is a violation of both the First Amendment speech right and the Fifth Amendment prohibition against a government taking of private property without “just compensation.” This week, the DC Circuit Court of Appeals agreed with Valancourt’s Fifth Amendment takings claim and declined to weigh the First Amendment claim.[1]

We agree with Valancourt that Section 407’s demand for physical copies of works, as applied by the Copyright Office here, represents an uncompensated taking of private property under the Takings Clause. We need not reach Valancourt’s First Amendment claim, as it seeks the same relief through that challenge.

Background (Summarized from Post of August 2021)

Valancourt produces books on-demand from old, rare, and hard-to-find manuscripts, including some dating back to the 18th century. Although many of the underlying works Valancourt republishes are long in the public domain, the companion material in their editions (e.g., scholarly introductions and footnotes) are subjects of copyright; and some of the manuscripts they print are still under copyright protection and used by permission of living authors or their estates. So, although the publisher does not register its books with the Copyright Office, it does place copyright notices in the front matter to alert readers that either the underlying manuscript and/or supporting materials are protected.

In 2018, Valancourt began receiving letters from the Copyright Acquisitions Division (CAD) of the Copyright Office demanding physical copies under the statutory deposit copy provisions of §407. The initial demand was for two copies of each of Valancourt’s 341 published titles. The publisher stated that this represented a significant financial and logistical burden and, finding no satisfactory relief in the Office’s compromise offers, they filed suit on the constitutional allegations. The district court ruled in favor of the government, and the DC Circuit has now reversed.

The Heart of the Decision

The most substantive aspect of the Valancourt decision is the court’s holding that the demand for physical copies is a taking because the publisher does not receive any benefit for supplying the copies. “A voluntary exchange for a benefit…does not exist if the purported ‘benefit’ is illusory,” the opinion states. Because copyright attaches automatically to any work fixed in a medium of expression, a party like Valancourt truly gets nothing in exchange for providing physical copies.

What the Valancourt decision means for most individual creators is not much, but it may be easy to confuse the issues here with the registration process (§408) in which deposit copies are a requirement of the registration application. Many creators are aware that in order to apply for copyright registration, they have to send in a copy of their work. That is governed by §408. In contrast, §407 allows the Library of Congress – through the Copyright Office – to demand a copy of any work published in the United States. Sending a copy for registration under §408 can satisfy the requirement under §407, but because Valancourt does not register the books it publishes, that was not relevant here. Consequently, the Office’s demand for physical copies (based solely on the act of publication) is acutely unjust in this instance because Valancourt gets absolutely nothing for providing free copies for the Library’s collection at its expense. This is distinctive from the deposit copies required for registration.

The deposit copy submitted for a registration application, whether electronic or physical, is less likely to be viewed as a taking because the applicant voluntarily obtains key enforcement advantages by complying with the registration requirements. Registration is not mandatory for copyright rights to subsist but is (among other things) a prerequisite to filing an infringement claim in federal court. Physical copies are required with a registration application if the work is published at the time of the application and if the work is first published in physical form, but again, because this is a condition of registration, and registration provides tangible benefits, this deposit condition is less likely to be found a taking.

That said, where the Office may exercise its right to demand copies based on publication (§407), the Valancourt decision may have some interesting implications. For example, if an author registers a final-edit manuscript in electronic form and later publishes that material, she (or the publisher) may receive a demand to provide physical copies or face a fine for failing to comply. But after Valancourt, the demand for physical copies solely at the threshold of publication, would seem to be a taking because, again, the author/publisher gets no additional value in exchange for providing the copies. The underlying registration (e.g., in the previously unpublished manuscript) is not invalidated by a failure to provide post-publication physical copies.

This is More About the Library of Congress than Copyright Law

In Valancourt, we see one example of the tension that can arise between the interests of the Library of Congress and the efficient administration of copyright protection for creators. As discussed in other posts, the often-confusing formalities in U.S. copyright law are largely due to the historic, but not always compatible, relationship between the two institutions. The Copyright Act of 1870 consolidated registration and deposit at the Library for the purpose of growing the collection, and in 1897, the Copyright Office was established within the Library as both the administrative and expert agency.

But as the outcome in Valancourt highlights, the author/publisher’s copyright interests and the Library’s authority to demand copies are in conflict. With 1870 barely visible in the rearview mirror, the Library still wants free materials, but Valancourt shows that this mission is not wholly predicated on the author/publisher’s interest in obtaining the full advantages of copyright protection. And perhaps this is especially true in a modern, digital-age market.

Notwithstanding the unpublished manuscript scenario above, most publishers send physical deposit copies with registration applications for published—or about to be published—books.[2] But if the Valancourt decision stands, it is conceivable that many publishers, especially smaller presses, will argue that they are allowed to submit electronic copies or must be compensated for the cost of sending physical copies. We shall see whether the government appeals this decision to the Supreme Court, but in the meantime, most creators can file this case under the heading Conversation for Copyright Nerds and continue registering their works with electronic deposits. In any event, please remember to register immediately—before your work is published and before it can be infringed – to get the full protection of the Copyright Act.

[1] FWIW, I wonder whether Valancourt’s speech claim would have merit because it seems that Plaintiff would have to show that the demand for physical copies, though enforced inconsistently, is content based. City of Austin v. Reagan National Advertising of Austin, LLC, 142 S. Ct. 1464 (2022).

[2] There are good reasons to do this, not the least being that the physical book is the final draft of all the material being registered.

Thaler Suit Against Copyright Office Asks for Analog Reading of Statute

Last February, the U.S. Copyright Office rejected the registration application filed by Stephen Thaler for a visual work entitled “A Recent Entrance to Paradise.” Thaler averred that the image was generated by an AI he designed called “Creativity Machine,” and on that basis, the Office affirmed the longstanding doctrine that copyright rights only attach to works of human authorship. In a series of cross motions, Thaler now argues that “non-humans have been authors under the statute for more than a hundred years.”

Broadly, Thaler asks the court for statutory interpretations based on proving a negative—namely that the Copyright Act does not explicitly state that an author must be human, or that an author cannot be non-human. His brief is peppered with examples in which the law does not expressly prohibit a non-human from having an “idea” or making creative “choices.” And he’s right. Nothing in the law makes such statements, but why would it? It is only very recently that law—and not just copyright law—must confront issues presented by sophisticated machines capable of performing functions ordinarily reserved for humans.

Prior to the present moment, it would have been absurd to affirmatively state that works of creativity or invention must be made by human beings. And it is frankly still absurd. Vehicular law does not explicitly state that the rules apply solely to human drivers (though we may have to address this one); marital law does not explicitly state that the parties must be human; and most relevant to this discussion, employment law does not explicitly state that non-discrimination and other rules of fairness apply only to human persons. Employment law is informative because Thaler’s primary claim of copyright in the visual work rests on the Work Made for Hire (WMFH) doctrine on the basis that “Creativity Machine” is the legal equivalent of an employee. But here, Thaler asks the court to read the law both strictly and metaphorically at the same time. Because nowhere in employment law are any rights vested in the analogous employee. For instance, the U.S. Equal Employment Opportunity Commission states:

Under the laws enforced by EEOC, it is illegal to discriminate against someone (applicant or employee) because of that person’s race, color, religion, sex (including gender identity, sexual orientation, and pregnancy), national origin, age (40 or older), disability or genetic information.

Title VII does not emphatically state that those qualities must describe human persons, but if we look to the definition of “persons,” we find …

The term “person” includes one or more individuals, governments, governmental agencies, political subdivisions, labor unions, partnerships, associations, corporations, legal representatives, mutual companies, joint-­stock companies, trusts, unincorporated organizations, trustees, trustees in cases under Title 11 [originally, bankruptcy], or receivers.

So, if we are following Thaler’s supposedly rigorous statutory interpretation, we can end the discussion by noting that there is no mention of computers, algorithms, machines, robots, etc. to support the argument that “Creativity Machine” is an employable “person” as an operation of law. Indeed, even the non-human entities in the definition above do not exist except as various means to organize the work or interests of human persons, and this is relevant to the copyright question presented.

While it is true that entities like corporations can own copyrights, the rights themselves do not exist until the moment of fixation of a human’s conception and the result of a human’s effort. Until then, there are no rights which may be transferred to the entity—either by independent or employment contract. Human authorship is not removed from the production of the work by the administrative formalities entailed with corporate ownership. On the contrary, human authorship must occur in order for the rights to exist at all.

This same principle applies to Thaler’s semantic game alleging that pseudonymous or anonymous works become works of non-human authorship merely because the term of protection is no longer based on the lifetime of the author. This makes no sense. Does Thaler argue that works made prior to the 1976 Act were not works of human authorship simply because they were protected under fixed terms unrelated to the life of the authors? He is conflating administrative rules (which are admittedly more complicated than necessary) with the doctrine of human authorship, which predates those rules by a few centuries.

Simply put, there is no concept of copyright law anywhere in western culture in which the utilitarian purpose of incentivizing authors to produce and disseminate works is not intertwined with the principle that the fruits of intellectual labor are the property of the author as a matter of natural right. The hyper-utilitarian view often overlooks the natural rights vested in the author, but the bundle of rights codified in Section 106 of the Copyright Act are as intangible and violable as any other civil right articulated by statute.  And such rights do not exist without humans.

Thaler and others are free to invoke philosophical debate as to whether an AI can have “ideas” or make creative “choices,” and it’s all fine chatter for a round of drinks, but even if it could be proven that the AI is “conscious enough” to make a choice to create and how to create, this is irrelevant as a matter of law. And we do not even need to confront AI per se. An elephant can make a painting, and a circus can own an elephant, but the circus cannot claim copyright in the painting as a WMFH because neither copyright rights nor an employment agreement exists between elephant and circus. And this is because no rights—copyright or otherwise—exist for the elephant that give her standing in court.

Personally, I believe the courts have gone too far in blurring the line between corporate and human personhood—e.g., vesting Hobby Lobby, Inc. with the right of religious exercise, but even in that controversial case, human persons, as owners of a closely held company, remain at the center of the finding. This is not so in Thaler’s claim of copyright in the visual work at issue. “Creativity Machine” is not a “person” by any statutory definition and cannot, therefore, have agreed to an employment contract in which it was understood that the employee would produce visual works under the WMFH doctrine. It is an absurdity on its face scratching at pencil shavings of statutory meaning, and Thaler’s appeal to history does him no favors.

I suspect the courts will find Thaler’s reading of the law to be untenable and will further advise that only Congress can change the Copyright Act. And if Congress were to do so in this context, I would hope that they affirmatively state that authorship must be human. If elephants don’t have such rights, why should robots?

Robot image by: sarah5