Pursuing Copyright Clarity Amidst The Gen AI Creativity Revolution – Copyright

Introduction

A collision is on the horizon. The collision is between a strict
interpretation of the human authorship requirement under U.S.
copyright law, and the ascendence of generative artificial
intelligence (Gen AI) as an essential element of modern creativity.
From publishing and advertising to music, film, media, and software
development, AI systems are reshaping workflows and redefining
authorship. Yet, as this technological revolution accelerates, U.S.
copyright law remains anchored to an interpretation that feels
increasingly out of step: that Gen AI outputs are all ineligible
for copyright and, thus, automatically ejected into the public
domain.

The U.S. Copyright Office has taken a firm position on this
point. Unfortunately, this stance, while perhaps doctrinally
consistent, is misaligned with economic reality. Our
copyright-reliant industries – arguably some of the crown
jewels of the U.S. economy — cannot tolerate prolonged
uncertainty about copyright ownership, protection, registration and
enforceability. If copyright law and practice fail to adapt, there
could be economic consequences for U.S. competitiveness. And, even
at the individual level, there is a growing sense that the law
should not deprive authors of the fruits of their genuine
authorship, even when intertwined with Gen AI.

Consideration should be given to the ways in which copyright
could evolve to provide greater certainty to our creative
industries, while remaining aligned with the foundational purposes
of copyright law. This article — in an
all-too-cursory-but-constrained-to-blog-post manner —
examines three possible paths forward, namely: (1) amending the
Copyright Act to construe Gen AI outputs as works-made-for-hire,
(2) amending the Copyright Act to allow Gen AI outputs to be viewed
as joint works with human users, and (3) allowing the law to
continue to evolve organically through the courts and agency
guidance. Each approach carries implications for creators,
businesses, and the integrity of copyright law.

The Human Authorship Requirement: A Doctrinal Anchor

Copyright law in the United States rests on the premise that the
copyright monopoly incentivizes human creativity. Section 102 of
the Copyright Act protects “original works of
authorship,” and courts have consistently interpreted
“authorship” as requiring a human creator. This principle
was reaffirmed in Thaler v. Perlmutter, where the DC
Circuit held recently that a Gen AI system cannot be an author
under current law, and that Stephen Thaler, the system creator and
user in the case, was also not the author of the work.

The Copyright Office’s guidance echoes this view: works
“produced by a machine or mere mechanical process that
operates randomly or automatically without any creative input or
intervention from a human author” are not registrable. The
rationale is understandable — copyright exists to reward
human creativity.

Yet, the boundary between tool and creator is blurring. When a
human conceptualizes the work and provides prompts, but the Gen AI
determines composition and detail, is this human authorship? If a
designer uses Gen AI to generate hundreds of variations and selects
one for refinement, does that selection, together with the
refinement, constitute sufficient creative input or control? At
what point does a combination of preliminary concepts, prompts,
selection and refinement cross over from non-authorship to
authentic, protectable human authorship? These questions expose the
inadequacy of a rigid test in an era of collaborative
creativity.

Specific Areas of Concern

The creative community appears to be coalescing around at least
three key areas of concern with the present state of play:

Exclusion of Authentic Authorship. There is a
sense that the Copyright Office’s decisions to date are
yielding inequitable denials of copyright to works that in fact do
constitute authentic human authorship. As now well-known, Jason
Allen undertook 624 prompts and significant editing and refinement
before finalizing his award-winning Théâtre
D’opéra Spatial; Kris Kashtanova developed 18 pages
of images displaying consistent tonal and thematic elements,
including significant editing and refinement by Kashtanova, in
creating their Zarya of the Dawn comic; and Stephen Thaler
programmed and trained his own Gen AI tool, and then directed its
output of A Recent Entrance to Paradise. Are these artists
really undeserving of copyright protection for these works? Do the
works not embody sufficient origination, conceptualization,
organization, selection, control, creativity, execution,
refinement, superintendence and stewardship to constitute human
authorship? Does denial of copyright protection for these works
serve the objective of promoting artistic progress? For many, the
answer to these questions is no.

Burdensome/Potentially Impossible Copyright
Registration Procedure. Another area of concern involves
the Copyright Office’s new requirement that applicants describe
and disclaim any more-than-de-minimis contributions to the
work created using Gen AI. Per the Copyright Registration
Guidance: Works Containing Material Generated by Artificial
Intelligence, applicants are asked to separate portions
created by the human author from portions created by the AI. In
many cases, this is an extremely burdensome and even potentially
impossible task – works often go through an iterative process
where dozens to hundreds of drafts are prepared, revised, further
revised and so forth, in order to achieve the final product. Under
those circumstances, can an applicant always accurately separate
which portions were created by AI and which by the human? And if
they get it wrong, is the registration subject to invalidation
(thus putting statutory damages, attorneys fees, and the other
benefits of registration at risk)?

Litigation Uncertainty. A final concern
involves copyright infringement litigation. Under current law,
virtually every defendant in every copyright infringement lawsuit
going forward will be motivated to challenge the validity of the
underlying copyright based on the potential presence of Gen AI
portions. Even if a plaintiff has complied with the law and only
claimed the human-authored portion of a work in the work’s
registration, the copyright could still be challenged. And, if the
plaintiff is unable to produce detailed records proving the exact
portions contributed by the human author versus the Gen AI, the
copyright could ultimately be found unenforceable. This creates a
tremendous burden and great uncertainty for the future for all
copyright infringement plaintiffs, and seems ultimately
untenable.

The remainder of this article considers three possible ways in
which copyright could evolve to meet the above challenges, to
ensure both the continued robustness of U.S. copyright law as well
as creating a welcoming environment for the increasing use of Gen
AI to advance innovation, creativity, and productivity in our
creative industries.

Works-Made-for-Hire?

One proposed solution would be to amend the Copyright Act to
allow AI-generated works initiated by a human user to qualify as
“works-made-for-hire.” Under current law,
works-made-for-hire include those created by employees within the
scope of employment or certain commissioned works under written
agreement. Extending this concept to the Gen AI context would vest
ownership of Gen AI outputs in the user, providing clarity and
predictability.1

The advantages of this approach are significant. It would offer
creative individuals and businesses copyright certainty, enabling
them to invest confidently in various types of Gen AI-driven
projects and use cases without fear of losing intellectual property
rights to the public domain. Questions would be reduced about
whether enough control was exerted by the human user, how
authorship should be identified on copyright applications, and how
copyrights could be enforced in copyright litigation. Any need to
document clear separation of the contribution of the Gen AI versus
the contribution of the human author would be mitigated.

This structure also makes some conceptual sense: when an
employer engages an employee to develop a creative solution
(whether in text, software, visual art, music, or otherwise), the
employer defines the problem, task, or concept, the employee
executes creative work to design a solution, and the parties may
then work together to further refine the solution until finished.
In that case, the employee would ordinarily own the copyright in
her creative work, but that result is flipped by the
work-made-for-hire doctrine – with the ownership and
authorship of the work vesting in the employer instead of the
employee. Analogously, a Gen AI user ideates a problem, task, or
concept, prompts the Gen AI system accordingly, the Gen AI executes
creative work to produce a solution, and the user may further
refine the output, either with or without the AI’s assistance,
until a satisfactory final product is achieved. Instead of the Gen
AI being viewed as the author, the ownership and authorship of the
work would vest automatically in the human user.

However, there are important drawbacks to this approach which
cannot be ignored. Gen AI systems are neither employees nor
contractors, and redefining “work-made-for-hire” to
include non-human entities may distort statutory coherence and
create other interpretive problems. This approach could also be
over-inclusive and dilute the authorship principle, incentivizing
minimal human input and undermining the creative threshold that
copyright seeks to protect and encourage. The question would
invariably be asked, do we really want to award a copyright
monopoly to a work fully created by Gen AI with, say, a single
three-word prompt by the human user? Is the objective of promoting
progress in the useful arts still being met with this solution?

Joint Works?

Another proposal is to treat Gen AI outputs as joint works, with
the human user as co-author. Joint authorship under current law
requires two or more authors intending their contributions to merge
into a unitary whole, and with each contributing sufficient
material to constitute authorship. Extending this concept to Gen AI
would acknowledge the collaborative nature of human-AI
creativity.

The advantages of this approach lie in its realism. Many
creative processes now involve iterative human-AI interaction, and
recognizing joint authorship would reflect this dynamic. It also
preserves human contribution, ensuring users retain rights even
when Gen AI plays an equal or even dominant role. Moreover, this
model offers flexibility, adapting to varying degrees of human
input and accommodating diverse creative workflows. It would also
still require non-minimal contribution by the human author, and
would allow for both the Gen AI and human portions of the work to
be safely protected by copyright and not exposed to potential
relegation to the public domain.

Yet, this solution is also not without challenges. Gen AI cannot
own anything, making “joint authorship” a legal fiction.
Further, determining the threshold of human input required for
authorship would introduce ambiguity. This approach would still
suffer from the need for the human to demonstrate a non-minimal
contribution, and courts would struggle to apply consistent
standards. This uncertainty could lead to litigation difficulties
that may be no better than the challenges of the present legal
situation, with disputes over how much human authorship was present
and whether this amount was sufficient for the human to be
considered a joint author of the work. (That said, the minimum
threshold for determining whether a contribution is sufficient for
participation in a joint work is well-trod legal terrain; in
theory, this would be a significantly better and more certain area
to litigate versus the as-yet uncertain area of identifying the
boundary between copyrightable human-authored works assisted by Gen
AI and Gen AI-produced works ineligible for copyright.)

Organic Development; Just Let it Be?

The third path is to let the law evolve organically through
judicial decisions and Copyright Office guidance. Under this
approach, courts would continue to distinguish works that are
“substantially Gen AI” (ineligible) from those with
“sufficient human authorship” (eligible), refining
standards case by case in a slow and laborious process.

This approach offers several advantages. It avoids premature
legislative overhaul, allowing nuanced, fact-specific evolution as
technology matures. Judicial flexibility ensures that courts can
adapt standards to emerging realities without locking in rigid
statutory definitions. Importantly, this path preserves core
principles, maintaining human authorship as a doctrinal anchor
while accommodating incremental change.

However, the drawbacks are considerable. Prolonged uncertainty
leaves businesses and individual creators navigating a legal gray
zone, exposing them to risk and inconsistent outcomes. This
ambiguity could chill investment in AI-driven creativity, deterring
innovation in sectors that depend on copyright certainty at a time
when these industries should be innovating and taking advantage of
the opportunities offered by AI augmentation. Fragmentation is
another concern, as different courts and agencies could adopt
divergent approaches, creating a patchwork of standards that
complicates compliance.

Conclusion and Next Steps

Unlike many other hype-curve developments in recent years, Gen
AI truly seems to challenge copyright law at a fundamental level.
Strict interpretation of the human authorship requirement to
exclude Gen AI outputs seems increasingly impractical in a world
where creativity is often co-produced by humans and machines. No
solution is perfect. Legislative reform offers clarity but risks
doctrinal distortion; organic development preserves principles but
prolongs uncertainty.

In the short term, the Copyright Office could refine its
guidance to provide clearer thresholds for human authorship,
reducing ambiguity for creators and businesses. It also could
consider acknowledgement of a broader array of contexts in which
human authorship may be recognized and potential relaxation of
application requirements. In the long term, Congress will likely
need to act — whether through a works-made-for-hire model, a
joint authorship framework, a hybrid approach or something else
— to ensure that copyright law remains fit for purpose in an
AI-driven economy and can continue to incentivize creative
expression.

This article covers complex topics in insufficient detail.
Further research is warranted in the following areas:

How do the philosophical foundations of our concept of
“authorship” inform our perception of human authorship in
the AI era? Is our “author concept” a relative historical
notion that could be updated to meet current practical reality, or
is it necessary for legal continuity?

What exactly does the Constitution’s mandate to promote
“progress” in the useful arts mean? What types of
behaviors and outcomes – aesthetically, economically,
culturally, and phenomenologically – are we intending to
promote with copyright?

Are our copyright-reliant creative industries really so
copyright dependent that AI disruption to copyright ownership
certainty poses a serious economic risk? Is it possible that newer
business models and technology are sufficient to render uncertainty
in copyright not so bad?

Some works, like the rejected Allen, Kashtanova and Thaler
works, seem worthy of copyright protection despite the role of Gen
AI, while others – an image, say, resulting from a three-word
prompt to Midjourney with nothing more, e.g. – do not. What
principles should guide the Copyright Office and law to better
determine which works are monopoly-worthy and which are not? Would
there be harm in granting at least “thin” protection to
users in connection with Gen AI outputs?

Footnote

1. As outlined in a prior post, Dr. Thaler, together with a
collection of scholars in an amicus brief, have advocated for the construal
of Gen AI works as works-made-for-hire, with Dr. Thaler as the
putative employer/commissioner and his Gen AI “Creativity
Machine” as the putative employee/contractor.

The content of this article is intended to provide a general
guide to the subject matter. Specialist advice should be sought
about your specific circumstances.