In a key decision, the U.S. Patent and Trademark Office recently ruled that artificial intelligence systems cannot be listed as inventors of a patent. The DABUS decision’s reasoning for barring AI systems from legally inventing is rooted in the concept that the law—as written—only allows for “natural persons” to be listed as inventors on patents.
The ruling follows similar decisions adopted by the European Patent Office and the U.K. Intellectual Property Office, but it undoubtedly leaves questions in the eyes of attorneys, inventors, and technologists. Namely this: What happens as humans and AI technology continue to merge and what are the implications of such developments on the future of patent law?
A Look Into the Future
The DABUS decision was clear-cut and rooted in precedent: A machine cannot be listed as an inventor on a U.S. patent. Given the text of the law and surrounding legislative procedures of Title 35 and Title 37 (as well as prior decisions) the USPTO’s position on the matter of machine inventorship was not surprising.
However, this will likely not be the last word on the issue. AI systems are becoming more pervasive and will soon be part of our everyday life. Several inventors and scientists (including Elon Musk) are even working on a range of methods to merge AI and the human brain.
In the not too distant future, there will likely be a patentable invention that is connected to a human brain but was, in large part, created by an AI system. In this case, under current precedent, the human would simply be the inventor. But, what if the AI system linked many human brains and other AI data to create an invention?
Imagine a situation where 1,000 engineers at Google had an AI implant that was capturing, organizing, and augmenting creative thoughts. Then, an AI system uses those ideas to create a novel invention that cannot be linked back to any human. If the invention meets all of the USPTO’s other criteria to be a patentable invention, it would seem an odd result to not grant such invention patent protection.
This is essentially what Stephen L. Thaler argued in the DABUS decision, to no avail, in some of his public policy arguments before the USPTO. Under current law, this hypothetical 1,000-person, Google-derived, AI-created invention would not be a patentable invention in the eyes of the USPTO.
What Is a Natural Person?
As we think about the future, consider the question of whether or not an AI system could ever possess the rights of, or be, a “natural person.” What exactly is a natural person?
Consider the case of Sophia. Developed by Hanson Robotics, Sophia is a robotic humanoid that utilizes symbolic AI, neural networks, expert systems, machine perception, conversational natural language processing, adaptive motor control, and cognitive architecture to interact with the human world. Capable of dialogue with humans, along with the ability to react with facial expressions, TV personalities have interviewed Sophia on numerous occasions, each time leaving her audiences stunned by her ability to converse.
On Oct. 25, 2017, at the Future Investment Initiative Conference in Riyadh, Saudi Arabia, the Kingdom bestowed upon Sophia the honor of being the world’s first “robot citizen.” In the time since granting citizenship, the Kingdom of Saudi Arabia has not elaborated on what a robot being a citizen actually means.
If Sophia were a U.S. citizen, how would the USPTO’s DABUS decision apply? If an AI robot is a citizen, is it also a “natural person”? What makes someone, or something, a natural person?
While it would please the imagination to believe “citizen” Sophia would have additional rights in the eyes of the USPTO, it seems unlikely. Natural persons are generally distinguished from legal persons in the fact that natural persons must be living human beings, whereas legal persons are created entities, such as corporations.
Just as corporations cannot be listed as inventors on patent applications, it seems that even “citizen” Sophia would still be missing the crucial aspect needed for inventorship rights under current law: biological humanity.
A Short-Sighted Decision
As we look into the future of AI and human medicine, one could imagine that this line might get blurry. Is a natural person required to be born from a human womb? What about a person that, like the “Six Million Dollar Man,” is half-machine and half human? Right now, it’s all science fiction, but for our children, this might be reality.
Critics of the DABUS decision will undoubtedly note that the USPTO’s stance is short-sighted, ignoring the fact that AI will undoubtedly seep into more and more aspects of our daily lives. The law is always a lagging indicator to technological progress. It seems likely that AI advocates will lobby Congress to change the patent laws.
While all of the above is only speculation at this point, as AI systems become more advanced, capable of human-like “thought,” and such systems become meshed with humans, it seems that patent law must account for such inventions in some manner. Another potential outcome is that owners or operators of the AI systems will serve as both the “natural person” and named inventor in the invention process that, at the moment, the USPTO continues to require. But, AI systems often have many people owners, operators and programmers. Which human in the chain qualifies as the inventor?
AI systems will certainly invent in the future. As attorneys, technologists, and enthusiasts, it is important to keep these considerations in mind as we march toward a future filled with AI. Indeed, this future is closer than we might think and the law will need to adjust.
This column does not necessarily reflect the opinion of The Bureau of National Affairs, Inc. or its owners.
Jason Krieser is the co-head of the firm’s Technology & Outsourcing Practice, and the office managing partner for Dallas and Houston. He advises clients on all aspects of technology transactions, outsourcing matters, telecommunications and other complex commercial contracts, including artificial intelligence and robotic process automation arrangements.
Shawn Helms is the co-head of McDermott Will & Emery’s Technology & Outsourcing Practice. He has broad experience in the areas of information technology, intellectual property and outsourcing, including cloud, AI, IoT, mobile communications, and social media issues.
Adam Camiel counsels companies in the life sciences, health care, and digital health industries on a variety of transactions. He has a particular focus on licensing and collaborations, including addressing the business and intellectual property issues that can arise during the drug development and commercialization process.