With the advent of Artificial Intelligence in every imaginable sphere that matters, it is hardly surprising that innovators are now in a rush to build their Intellectual Property portfolio in AI. The popularity of Artificial Intelligence, machine learning, neural networks and all other things related to Intelligent machines has been on the rise for quite some time now. A similarly whopping growth has been observed in the number of machine learning or AI related patents/applications which have been filed in the recent years. A recently published article revealed that patent applications based on machine learning have nearly quadrupled in the last few years, owing to the widespread use of the same in all major technological fields. Since the formal conception of Artificial Intelligence and machine learning, way back in the 50s, more than 300,000 patent applications and over a million scientific journals have been published.
Further, within the elaborate world of AI, machine learning has always been a defining aspect which attracted significant attention from the very beginning. WIPO has stated that over one-third of the published patents/applications in the AI domain are related to Machine Learning. It has been estimated that the average annual growth in patents related to machine-learning is somewhere between 25-30 percent. Machine learning techniques such as deep learning, reinforcement learning, neural networks have seen unprecedented growth in terms of patent filings in the last few years. So, here’s a question – What does this flourish of activity reveal about the future of artificial intelligence in the world of Intellectual Property? The answer, as it turns out, is many-fold.
AI has often been compared to the “Electricity of the new era”. If so, Machine learning can surely be considered to be the wire which conducts it. WIPO has recently taken upon itself to develop a framework for the field of AI, including the applications of machine learning such as processing of speech and digital signal processing. The vital areas which would be completely transformed by AI were also identified, such as Transport and Communication. However, other sectors such as Law, Banking, Medicine, Entertainment, Agriculture are not far behind either in utilizing machine learning and AI, and building their respective patent portfolios as well. Bearing these facts in mind, it is not hard to see the growing interest of inventors in patenting their AI end applications.
On a global level – Microsoft, IBM and the like have been working tirelessly to build a hefty patent portfolio in the field of AI applications. The recent patent activity of IBM and Microsoft clearly suggest that these companies want to make their mark as pioneers in AI and machine learning as well. Toshiba, NEC and Samsung are major players in this race too, however, it will be a while before they can catch up to either IBM or Microsoft. A particular focus in recent times seems to be on machine learning techniques which are based on biologically inspired methods and supporting vector machines using supervised learning.
Being one of the fastest growing economies in the world, India has also been showing a significant growth in filing patent rleated to AI. Some recent studies show that India is emerging as a new target for patent filing in AI and is among the top countries for publications in specific categories such as computer vision and natural language processing. Reports on worldwide filing trend for AI rank India at the eighth place for first filings in 2015. Patent filing in AI in India has enjoyed a healthy growth rate in recent years (with an average of 33% up to 2015). Specifically,for AI functional applications, India holds fifth rank in patent filingsrelating to distributed AI. According to a patent survey conducted by us, Microsoft and Accenture are amongst the top patent filers in India. Some indian origin companies that make it to the list of top patent filers are Tata, and HCL technologies.
All in all, the AI seems to be poised to push the filings further up in coming times. However, it remains to be seen how the patent office will react to these filings – what objections to such applications will be made while examination. How difficult will it become to get a granted patent in the domain, especially when a lot of literature is already available in form of scientific publications, and expired patents since years.
By Mr Amit Aggarwal, Co-Founder and Director Effectual Services