Are Digital Inventors Real? Issues with Artificial Intelligence and Patent Invention
Artificial Intelligence (“AI”) is becoming central to today’s economy. From chatbots and healthcare services like machine learning systems, to various applications of creating useful information, AI has been increasingly used to optimize creative progress.[1]
However, there are several questions about AI and intellectual property protection. For example, how will current protections recognize the fruits of the creative and inventive processes of AI? Should they be given the same protection as inventions and works of human inventors and get ownership rights over an invention? A recent case addresses this line of questions. On September 3, 2021, the U.S. District Court for the Eastern District of Virginia held AI systems do not qualify as an “inventor” under the Patent Act, which otherwise requires an inventor be a natural person.[2]
In 2018, Stephen Thaler filed a patent application on behalf of DABUS, an AI “creativity machine.”[3] The USPTO rejected the application, arguing a machine could not qualify as an inventor.[4]. After a motion for reconsideration came to the same conclusion, Thaler filed this lawsuit.[5] Thaler argued below that a patent application should not be rejected only on the ground that there is no natural person identified as the inventor, specifically arguing no rejection when an AI otherwise fulfills the criteria for inventorship.[6] The government defines a patent inventor in many ways, but one is important: “The threshold question in determining inventorship is who conceived the invention. Unless a person contributes to the conception of the invention, he is not an inventor.”[7] A.I. systems also have many definitions, including the computerized simulation of human thought processes.[8] Put another way, machines are taught to operate like humans think, which theoretically could make them a person.[9] Therefore, in this case, the question was whether a patent application should be rejected on the ground that the inventor was not a natural person.
Thaler encouraged the court to interpret inventorship in a way “consistent with the Founders intent in enacting the Patent and Copyright Clause”.[10] According to Thaler, this interpretation would enable “to promote disclosure of information and commercialization of new technologies.”[11] Thaler also noted “[D]enying patent protection for AI-Generated Inventions threatens to undermine the patent system by failing to encourage the production of socially valuable inventions.”[12] According to his point of view, since AI inventions may one day become the primary source of innovation, giving such a literal meaning to the word “inventor” would not align with promoting new technologies.[13]
However, the court disregarded these arguments, stating that they were mostly speculative and based on policy considerations rather than law.[14] Moreover, the court stated that in 2011, Congress enacted the American Invents Act, which formally amended the Patent Act to provide an explicit statutory definition for the term “inventor” to mean “the individual.”[15] Their reasoning consisted in saying that if Congress had intended AI systems to be considered as “inventors” they would have explicitly included AI systems in the definition since in 2011 AI system already existed.[16]
Furthermore, in its analysis, the court made a parallel with the word “individual” used in the Torture Victim Protection Act (“TVPA”) which was interpreted as meaning “natural person” and more precisely “human being”.[17] By applying this interpretation to the case at hand, the court found that AI systems could not be characterized as inventors under the US Patent Act.[18].
This decision underlines one of many issues that AI may trigger regarding patent and other intellectual property matters – who is really the “author” whose work is protected. At the same time, this case does not solve the problem at hand: how AI systems may obtain protection for their inventions? The court is still not clearly answering this question.[19]
Moreover, the question raised by Thaler in this case seems to divide courts around the world. Indeed, Australia and South Africa have recently already allowed AI system to qualify as “inventors” under their national patent law.[20] Thaler already tried to patent his AI system outside the US, but the European Patent Office twice rejected the application in 2019 and 2020.[21] The office noticed the AI system lacked legal personality and therefore could not meet the applications requirement of providing a first name, last name, and address.[22] It seems like the UK followed this approach since on September 21, 2021, the UK Court of Appeal ruled as well that an AI system could not be listed as an inventor on a patent application[23] These conflicting interpretations among countries are likely to pose significant problems in terms of legal security.
At a fundamental level, the question that should be addressed in the first place is whether, inventions created without human input should merit patent protection.[24] Since patents trigger important costs, it seems relevant to examine whether they can be justified by benefits that are only made possible because of patent protection.[25] Therefore, with this perspective, it may be possible to hope that Congress may one day decide to expand the scope of patent law, considering that AI system’s invention are so essential that they deserve to get their own patent protection.
Footnotes
