Libertatem Magazine


Contents of this Page

Artificial Intelligence in the World of IP

Artificial intelligence (AI) has emerged as a new frontier with far-reaching implications for the global community. It will have massive technological, economic, and social ramifications, transforming the way people function, live, manufacture, and distribute goods and services. While it is too soon to say, it is clear that artificial intelligence (AI) will affect conventional intellectual property (IP) concepts. 


The intellectual property (IP) system’s fundamental objectives have always been to promote new inventions and innovative works, as well as to provide a viable economic foundation for innovation and creation. If other goals of the IP scheme, such as “just award” and moral rights, are ignored, there is no need not to use IP to reward AI-generated innovations or creations from a strictly financial standpoint.

Patents, prototypes, literary and artistic works can all be described by the widespread use of AI technologies. For example, the life sciences produce enormous amounts of data that are valuable but do not qualify as inventions in the traditional sense. But first, they must sort out the rights and responsibilities that come with them. Many argue that, since data is the backbone of AI, it should be easily accessible to allow AI and other applications to be developed.

Data and algorithms create some basic IP issues, such as how to create property rights in a constantly evolving algorithm.

Demand for intellectual property rights continues to outpace global economic growth. The IP scheme, as it is called, is not going out of style anytime soon. It is being used more often than ever before. However, new problems are arising, and instead of replacing the current system, an additional layer of IP may be required. 


The other question is whether AI will own or infringe on intellectual property. Who would be held liable if an AI machine generates subject matter that infringes on third-party intellectual property? The problem here is that copyright infringement necessitates actual copying. The author of the infringing work had to have had access to the copyrighted work. The difficulty of proving that the infringer had access to the protected work could be much easier to solve in the case of an AI device that is allowed access to anything on the internet. 


In the future, AI systems will play an increasingly important role in IP administration. Given the high costs of gathering and cleaning vast amounts of data to feed AI systems, resource sharing should be encouraged.

“The international IP group should work together in the future to achieve high levels of interoperability when cost-effectively implementing AI-powered systems,” says Francis Gurry, Director-General of the World Intellectual Property Organization (WIPO).

WIPO is looking at ways to use training data provided by member states and other institutional partners to improve AI applications. In exchange, it provides the partners with new AI applications created using the data.

WIPO, for example, has created WIPO Translate, a state-of-the-art neural machine translation device powered by AI,” says Gurry. “This tool is being shared with 14 intergovernmental organizations and different patent offices around the world.”

Since the framework is reliant on data access and availability, designing such tools will help this sector improve by using and supplying data. The more fundamental question is how AI can alter the categories and principles of intellectual property. However, it is taking place at a time when, for the first time in 70 years, the world is devoting less attention to multilateral rule-making.

Since IP is an international phenomenon, it is a serious problem that extends beyond IP and must be addressed in this field. Technology, like the patent data it generates, is global. Patents are seldom associated with a single jurisdiction. As a result, global solutions with functional interoperability are needed. 




Building AI capability across IP offices is a major undertaking. Although AI has been around for a while, it has only recently emerged as a viable technical solution. The number of professionals with the necessary training and knowledge is small in this field, making it difficult to build in-house AI capacity, particularly in the face of competition from better-resourced, higher-paying private enterprises. Smaller IP offices are bound to face certain difficulties. AI systems rely on data and algorithms, and smaller offices have fewer of these. This creates a volume imperative, which forces the creation and implementation of AI applications in larger offices. In smaller offices, where the importance of applications is still manageable, it is less efficient. Open access to data associated with IP registrations for patents, trademarks, and designs is a widely established policy in the IP world. It would benefit smaller IP offices, which would be able to access these data in theory.  To overcome these obstacles, a greater focus on integration and teamwork will be needed.










About the Author