Artificial Intelligence refers to the creation and development of complex, human-like brains. It is a branch of science and technology that’s fairly new, often still considered as a part of science fiction. But with the advancement of technology and human effort, it has become a reality, at least in some aspects and variations. It is now an essential tool used in scientific data evaluation and computation. AIs use algorithms for generating potential solutions, which requires conducting various research experiments. The results are free from certain kinds of bias (certain kinds, because human intervention is still necessary to define targets and to set parameters for measuring success), as well as time constraints imposed by human limitations.
The state of AIs today is very far off from science fiction stories of fully realized and sentient electronic beings. However, it’s evident that AI capabilities are increasing over time. One of the areas where this can be seen is AI innovation. In earlier times, AIs served as data aggregators or catalysts for new innovations, but we’ve now reached a point where scientists envision autonomously created patentable inventions in the near, possibly immediate, future.
With the emergence of such advances in science and technology, there are essential questions regarding the legal implications of protecting the results of such systems. There is uncertainty about whether the present patent legal framework is equipped to accommodate computer-generated inventions.
Patentability of AI Inventions
Patent Laws are essential in revitalizing R&D investment by giving inventors a reasonable certainty of returns on the resulting inventions. Uncertainty concerning the validity and enforceability of a patent can weaken its market value and impact a right-holder’s ability to extract value through licensing or litigation. This goes against the prima-facie objective of the patent to accelerate research and innovation and stimulate knowledge diffusion.
A patent is a term for a license protecting scientific invention by providing exclusive right granted for any product or process that provides a new way of doing something, or offers a new technical solution to a problem. Patentability criteria may vary slightly with the law of the land. However, under International Legal Agreement TRIPS (Trade-related Intellectual Property Rights), to claim patent protection, an invention must be new, involve an inventive step, and be capable of industrial application. It is applicable in the case of any product, process, and in all fields of technology.
In the case of Mayo Collaborative Servs v. Prometheus Lab, the Supreme Court of the United States explained that AI is the essential tool for scientific and technological-related work, and creating monopolies on these through patents would obstruct and hinder innovation. One of the more significant concerns of AI is the possibility of it replacing human employment, to great economic detriment. It has been stated that patents should not be granted on claims which are a mere replication of human activity and do not involve any inventive step. Monopolies on these through patents would obstruct and hinder innovation, according to the U.S. Supreme Court.
Liability Issues of Patent Infringements by AIs
One of the main arguments for disallowing AIs to hold inventions lies in the difficulty of enforcing potential infringement claims that may arise in the future. A patent grants the inventor the sole right to use and sell the invention. When someone without authority uses, deals, or proposes to sell an invention, they are subject to liability for infringement. In the event of a violation, the infringer must pay damages to compensate for the loss. Failure to hold someone accountable for patent infringement encourages the use of AI, an entity that has not yet received legal recognition of personhood, for infringement.
One option could be an insurance system in which funds are created for paying infringement damages. If a human agent is held responsible for the infringement, the liability should be in proportion to the extent of authority delegated to the AI. This would require granting the status of legal person to the AI.
Recent Cases on AI and Patentability
The United States Patent Office (USPTO) issued a judgment in April 2020about application number Appl. No. 16/524,350, in which there was no human inventor or co-inventor, but rather the application listed an AI system “DABUS” (“Device for the Autonomous Booting straining of Unified Sentience”) as its inventor. Despite meeting all of the patentability standards, such as usefulness, innovation, and industrial applicability, the application got refused because there was no human inventor, as stated in the U.S. patent law, which states that only natural individuals may become inventors.
The Patent application for DABUS was filed in September of 2019 under the Patent Cooperation Treaty. In July of this year, the Companies and Intellectual Property Commission of South Africa awarded the world’s first Patent and authorized the AI mentioned above creator as its inventor. The idea includes interlocking food containers, especially beverage containers that are exceedingly easy for robots to grip. Many people have argued that the Patent is meaningless because its creator is an AI. Others have argued that this innovation marks a milestone moment in the history of patent law.
DABUS is characterized as a connection AI system. It operates based on machine learning, which has recently acquired popularity in these Covid times. The technology mimics brainstorming and generates novel inventions. DABUS is a type of AI that is referred to as a creative machine. Machine learning is the process through which innovative devices process and analyze information. Once the machine learning phase is complete, it will function independently without the need for human interaction.
There is currently no legislation that recognizes an AI system as an inventor. Every legislation in every country across the world follows the idea that an inventor must be a natural person. The position of AI has evolved dramatically since its inception, and it is now more of a science than a work of fiction, as it was previously, with robots and computers replacing humans. Artificial Intelligence is a machine’s capacity to simulate intelligent human behavior. With the advancement of technology, the role of AIs in our daily lives has grown significantly. But the law has yet to catch up- recognition of AIs and the effects it has on technology development have not yet been considered in a serious capacity.
There is a good chance in the near future of there being a flood of patent applications since AI technology is far more efficient than humans are at developing ideas. AIs are likely to radically change the frontiers of human invention, and the law as it is now is a ticking bomb of innovations possibly going unprotected. There needs to be more clarity, and a global consensus, to resolve this matter, or at the very least address it, as quickly as possible.
Author: Aranya Nath, Legal Intern at PA Legal.
In case of any queries, kindly contact us here.