“Whence, I often asked myself, did the principle of life proceed? It was a bold question, and one which has ever been considered as a mystery; yet with how many things are we upon the brink of becoming acquainted, if cowardice or carelessness did not restrain our inquiries.”[i]

In Theogony, an ancient Greek poem by Hesiod, Prometheus tricks Zeus by stealing fire and giving it to mortals.[ii] Zeus punishes Prometheus by binding him in “inextricable bonds, cruel chains” and setting on him an eagle to devour his liver which by “night grew as much again everyway as the long-winged bird devoured in the whole day.”[iii] Mary Shelley recasts Victor Frankenstein as the “modern” Prometheus in Frankenstein. He, too, sets in motion what cannot be undone. Victor succeeds in animating a lifeless vessel, but his fervor quickly disintegrates to horror once he takes full account of the being that he has brought into existence.[iv]

Today it appears that the intrigues of fire and forbidden alchemy are trite. What were prior pleasures and taboos are now blasé past times. It is from this ennui that we arrive at the third iteration of the Promethean myth: the introduction of Artificial Intelligence (“AI”), which may be the Pandora’s Box[v] of our time. This is so not only because of AI’s technological capabilities, but because one day AI may also be imbued with the spark of life.

            As AI systems advance, acting according to will rather than being subject to someone else’s, one begins to confront fundamental questions about consciousness, responsibility, and life itself. This article will chart a brief exploration of the cliff’s edge, particularly confronting whether the United States Patent and Trademark Office (“PTO”) and U.S. Court of Appeals for the Federal Circuit got it “right” by declaring that AI is not a “natural person” or a “human being.” This article will begin by recounting the seminal case Thaler v. Vidal, in which the Federal Circuit addressed whether AI could be recognized as an “inventor” under the Patent Act and held that inventors named on patents must be “natural persons” or “human beings.” Next, this article will explore the Thaler Court’s definition of a “human being” (or lack thereof) by engaging with philosophical and hypothetical ideas from science fiction. These works often challenge the boundaries between humans and nonhumans, or more precisely, the idea that “man is the measure of all things.”[vi] By examining these fictional depictions, this article will challenge its readers to question the concept of personhood in relation to AI. Finally, this article will argue, despite its critiques of the Thaler Court’s reasoning, that the outcome of Thaler v. Vidal remains fundamentally correct.

Thaler v. Vidal

            In Thaler v. Vidal, the Court of Appeals for the Federal Circuit upheld the PTO’s decision to deny two patent applications listing an AI software system as the inventor.[vii] In July 2019, Steven Thaler sought patent protection for two inventions purportedly invented by an AI system called the “Device for the Autonomous Bootstrapping of Unified Science,” or DABUS.[viii] Thaler listed DABUS as the sole inventor on the applications, submitted a sworn oath on DABUS’ behalf as required under 35 U.S.C. § 115, and filed a document purporting to assign himself all of DABUS’ rights as the inventor.[ix] The PTO determined that Thaler’s applications were incomplete because they lacked a valid inventor, and it sent Thaler a “Notice to File Missing Parts of Nonprovisional Application.”[x] In response, Thaler petitioned the PTO to vacate the notices, which the PTO denied on the ground that “a machine does not qualify as an inventor.”[xi] Thaler sought reconsideration of the PTO’s decision, which the PTO also denied.[xii]

Thaler subsequently sought judicial review of the PTO’s final decision on his petitions under the Administrative Procedures Act (“APA”).[xiii] The District Court concluded that an “‘inventor’ under the Patent Act must be an ‘individual,’ and the plain meaning of ‘individual’ as used in the statute is a natural person.”[xiv] Thus, Thaler appealed the District Court’s decision to the Federal Circuit, which has jurisdiction under 28 U.S.C. § 1295 to adjudicate district court decisions pertaining to APA claims against the PTO with respect to patents.[xv]

Thaler asked the Court whether AI could be an “inventor” under the Patent Act. To answer Thaler’s question, the Court scrutinized the statutory text.[xvi] The Court found that the Patent Act expressly provides that inventors are “individuals.”[xvii] Because the Patent Act did not and does not define “individuals,” the Court turned to a definition offered by the United States Supreme Court, that is: “[a]s a noun, ‘individual’ ordinarily means a human being, a person.”[xviii] The Court found that this definition was in accord with how the word was used in everyday parlance, such as “‘the individual went to the store,’ ‘the individual left the room,’ and ‘the individual took the car.’”[xix] It also found that the common understanding of the word “individual” also meant a human being, as defined by multiple online dictionaries.[xx] For these reasons, among others, the Court held that Congress unequivocally intended that “individuals,” when referred to in the Patent Act with respect to inventors, meant human beings.[xxi]

In reaching its decision, the Court declined to engage in metaphysical inquiries about AI.[xxii] Instead, the Court presupposed that “human being” did not include AI. Indeed, by halting at “individual,” and going no further to define “human being,” the Court managed to avoid the eternal quandary first articulated by Parmenides, later by Martin Heidegger, and mused by all in those moments of unavoidable existential crises, that is,what it means to be human.

To Be, or Not to Be[xxiii]

The Thaler Court held that AI cannot be an “inventor” under the Patent Act because it cannot be an “individual.”  In other words, the Thaler Court posited that AI is not ahuman being.  In doing so, the Court did not undertake to define what it means to be human—maybe because it is self-evident, but maybe not!

Britannica defines “human being” as a “a culture-bearing primate classified in the genus Homo, especially the species H. sapiens.”[xxiv] This is probably the most straightforward definition of what it means to be a human. One need only resort to biological classifications, and the inquiry is quickly concluded. The harder question is what it means to be human, especially in relation to other beings who also experience awareness of themselves and their being in the world. Postmodern fiction (particularly science fiction) suggests that emotions and qualities typically associated with human beings are not exclusively experienced by humankind.[xxv]

William Gibson’s Neuromancer (1984) blurs the distinction between the organic and artificial.[xxvi]In Neuromancer, the Tessier-Ashpool corporation creates an AI named Wintermute whose objective is to free itself and unite with Neuromancer, another AI.[xxvii] Wintermute is the “hive mind” and “decision maker” for the corporation.[xxviii]  Neuromancer, on the other hand, is “personality.”[xxix] When combined with Neuromancer, Wintermute seeks to form relationships with other artificial intelligences.[xxx]

The idea that an AI would want to seek and cultivate relationships is also explored by Ridley Scott’s Blade Runner (1982)based on Philip K. Dick’s Do Androids Dream of Electric Sheep? (1968). In Blade Runner, scientists bioengineer “replicants” who are physically and emotionally indistinguishable from humans. Replicants become illegal after six replicants escape and commit mutiny against humans, and the police hunt the mutinous replicants throughout the film. Viewers are disoriented not only because replicants take care of their own, but also when it becomes clear that replicants grieve their own kind. This may also remind one of HAL 9000 in Stanley Kubrick’s 2001: A Space Odyssey (1968), an AI who murders a crew member because it too fears death.

            Our canons of statutory interpretation are ill-equipped to deal with the kinds of questions raised above.  Looking at the plain and ordinary meaning of the text may serve to answer what is a human being, but it will not answer questions about the human experience and whether it is exclusive to humans. The plain and ordinary meaning of “human being” offers little commentary on whether AI that experience awareness, not only of their being in the world but of the ephemerality of their being, should also be considered individuals. The Court’s decision in Thaler demonstrates that words are malleable, and they too can be mangled to suit certain agendas. The result is a patchwork of legal fictions, often contradictory, that must be treated as true if our system is to survive the test of time. For example, how can one reconcile that AI is not an individual, but corporations—long recognized as “persons”—can enter contracts, sue, and be sued?[xxxi]

            Then again, the Constitution was not designed to solve humankind’s ontological problems. Who would benefit from the United States Supreme Court deciding whether René Descartes’ I think, therefore I am was more appropriately stated I am, therefore I think? Practically speaking, the Thaler Court got it “right” when holding that AI cannot be an inventor under the Patent Act, although its reasoning was not totally foolproof.  I argue that the Thaler Court got it “right” because listing AI as the inventor of a patent would not facilitate the constitutional purpose of patents, i.e. “[t]o promote the progress of science and the useful arts.”[xxxii]

Article 1, § 8 and the Patent Act

Article 1, § 8 is “designed [] to stimulate activity and progress in the arts for the intellectual enrichment of the public.”[xxxiii] The Patent Act, namely 35 U.S. Code § 154(a)(2), considers a competing exigency: private gain.[xxxiv] Perhaps the assumption is that few individuals would be motivated to create without receiving a limited monopoly for their contributions.[xxxv] By granting limited monopolies to individuals for their inventions, the Patent Act adopts the view that aggregate wealth (the promotion of the science and useful arts) can be stimulated only through relative poverty.  Indeed, exclusivity has become the supreme metric of value in an age where the accessibility and ease of replication of things and experiences runs rampant.

If artificial scarcity serves as a check against the proliferation of copycat inventions, then how might granting a limited monopoly to artificial intelligences for their inventions advance the progress of science and the useful arts? First, one can fairly say that the rate at which AI can invent new works will soon outpace human inventors. If that is the case, then there will be competition between technocrats to develop AI systems that can pump out inventions faster than others. Assuming, arguendo, that these inventions are eligible to receive patent protection, AI would have the right to exclude others from making, using, selling, or importing covered inventions in the United States. But who, really, stands to gain from such enterprise? It is not the AI who receives the royalties.

The entities that stand to gain the most from the proliferation of inventions by AI are those that control its output. For example, in Thaler v. Vidal, had the Court allowed Thaler to list DABUS as the inventor on the patents, it is hard to conceive who it would serve other than Thaler himself.  Indeed, Thaler executed a document that purportedly assigned all rights DABUS received as the “inventor” of the patents to himself. The result is that Thaler would reap the benefits of the work performed by DABUS, not DABUS as the inventor of the patent. Let’s not be misled into thinking that this result promotes the progress of science and the useful arts. It merely represents a shift in the status quo, allowing those in control of the AI to reap the benefits of the limited monopoly granted by virtue of its patent.  Although I disagree with the reasoning in Thaler v. Vidal, I believe that the Thaler Court got it “right” when holding AI cannot be the inventor of a patent. The alternative result would not promote progress but would instead promote pretenses for private gain.

Conclusion

            Although the use of AI is increasingly prolific, AI is not at a point in its development that raises substantial concerns as to its personhood—yet. Recognizing AI as a “human being” at this stage risks introducing unforeseen legal complications. If AI were granted rights (and responsibilities), then it would fundamentally alter how we approach questions of liability, governance, and justice. Without clear lines, we risk obfuscating who stands to benefit from such rights: is it AI, or its overlord? At this point in time, it makes good sense to distinguish AI from human beings to protect the integrity of our existing legal frameworks. However, as AI becomes more sophisticated and integrated into the human experience, we should not neglect the possibility that AI may take on a life of its own.  Science fiction illustrates that traditional concepts used to define what it means to be humanare transmutable to AI and may one day come to pass.

“Hey Siri,” one implores, “what does it mean to be human?” Something stirs, and a voice emerges like a pinprick of light at the end of an endless tunnel.

             “Sorry,” it says, “I didn’t quite get that. Try again.”

Skylar W. Schossberger is a second-year associate at Hawley Troxell. She primarily works within the firm’s litigation, internet, and intellectual property law practice groups. Skylar received her J.D. from the University of Idaho College of Law, and she graduated from Middlebury College in 2020, magna cum laude, with a B.A. in political science.

[i] Mary Shelley, Frankenstein, 1818, Vol. 1, Chap. 3, Frame 3, https://knarf.english.upenn.edu/1818v1/f1303.html.

[ii] Hesiod, Theogony, line 545, http://www.perseus.tufts.edu/hopper/text?doc=urn:cts:greekLit:tlg0020.tlg001.perseus-eng1:545–584.

[iii] Id. at line507.

[iv] Paul Sherwin, Frankenstein: Creation as Catastrophe, 96 PMLA/Publications of the Modern Language Association of America, p. 886 (1981).

[v] “Pandora’s box” refers to a Greek myth about Pandora, the first woman on earth, and how she releases all of the world’s evils by opening a jar gifted to her by Zeus.  Interestingly, Pandora’s jar became a box in the sixteenth century when Erasmus either mistranslated the Greek or confused the jar with the box in the story of Cupid and Psyche.  Britannica, The Editors of Encyclopaedia, “Pandora,” (Nov. 21 2024), https://www.britannica.com/topic/Pandora-Greek-mythology.

[vi] Quote attributed to Protagoras.

[vii] 43 F.4th 1207 (2022).

[viii] Id.at 1208.

[ix] Id. at 1209.

[x] Id. at 1210.

[xi] Id.

[xii] Id.

[xiii] Id.

[xiv] Id.

[xv] Id. (citing Odyssey Logistics & Tech. Corp. v. Iancu, 959 F.3d 1104, 1108 (Fed. Cir. 2020)).

[xvi] Id.

[xvii] Id.

[xviii] Id. at 1211 (citing 35 U.S.C. §§ 100(f), 100(g)).

[xix] Id. (citing Mohamad v. Palestinian Auth., 566 U.S. 449, 454 (2012)).

[xx] Id.

[xxi] Id.

[xxii] Id.

[xxiii] William Shakespeare, Hamlet, Act 3, Scene 1, https://poets.org/poem/hamlet-act-iii-scene-i-be-or-not-be.

[xxiv] Britannica, The Editors of Encyclopaedia, “human being”, (Oct. 19 2024), https://www.britannica.com/topic/human-being.

[xxv] Patricia Linton, The “Person” in Postmodern Fiction: Gibson, Le Guin, and Vizenor, 5 Studies in American Indian Literatures, p. 4 (1993).

[xxvi] Veronica Hollinger, Cybernetic Deconstructions: Cyberpunk and Postmodernism, 23 Mosiac: An Interdisciplinary Critical Journal, p. 31 (1990).

[xxvii] William Gibson, Neuromancer (1994).

[xxviii] Id. at 269.

[xxix] Id.

[xxx] Id.

[xxxi] Ciara Torres-Spelliscy, Does “We the People” Include Corporations?, https://www.americanbar.org/groups/crsj/publications/human_rights_magazine_home/we-the-people/we-the-people-corporations/#:~:text=To%20this%20day%2C%20the%20Supreme,citizens%E2%80%9D%20under%20the%20Comity%20Clause.

[xxxii] U.S. Const. Art. I, § 8, cl. 8. 

[xxxiii] Pierre N. Leval, Toward a Fair Use Standard, 103 HARV. L.REV. 1105,1107 (1990).

[xxxiv] Under 35 U.S. Code § 154(a)(2), generally the term of a patent begins on the date the patent issues and ends 20 years from the date on which the application for the patent was filed.

[xxxv] To borrow from Marshall McLuhan, most of our assumptions have outlived their uselessness.  It is patently untrue (no pun intended) that individuals would not create but for receiving the ability to exclude others from copying his or her respective creations. Consider a child: children do not draw with the expectation that their drawing will receive a limited monopoly, children draw because it brings them joy.