Recent Artificial Intelligence Patent Trends
By: Katie Rubino
Artificial intelligence (“AI”) has transformed our world by becoming a ubiquitous presence integrated into many facets of everyday life. Today, chatbots effortlessly open and answer any questions we may have while deep learning algorithms aid to prevent cybersecurity breaches at airports. Recently, the United States Patent and Trademark Office (“USPTO”) released a report highlighting patenting trends in this highly technical space.
What is AI?
The first question released by the USPTO was to provide a definition as to what exactly constitutes AI. The USPTO adopts the definition set forth by the U.S. National Institute of Standards and Technology (“NIST”); they define AI to “comprise software and/or hardware that can learn to solve complex problems, make predictions or undertake tasks that require human-like sensing (such as vision, speech, and touch), perception, cognition, planning, learning, communication, or physical action.” This definition utilized by the USPTO includes eight various applications and subsets of AI—speech applications, knowledge processing, AI hardware, evolutionary computation, natural language processing, machine learning, vision, and planning and control.
AI Patent Activity
A review of patenting activity from 1976 through 2018 indicates a steady and significant increase in activity. From 2002 through 2018, annual AI patent applications increased by more than 100%, rising from 30,000 applications to more than 60,000 applications. The share of AI applications adjusted based on an overall trend reporting an increase in patenting activity showed an explosion of growth in the AI landscape, from 9% in 2002 to 16% in 2018. Of the eight various subsets utilized by the USPTO in the definition of AI, the sub-sets with the largest increase in patent activity included planning and control and knowledge processing.
From 1976 through 2018, the five U.S. companies with the most AI patent filings included: IBM Corporation with 46,752 granted patents, Microsoft Corporation with 22,067 granted patents, Google Incorporated with 10,928 granted patents, Hewlett-Packard Corporation with 7,072 granted patents, and Intel Corporation with 7,021 granted patents.
Top 10 U.S. AI Patent Owners
Geographical location throughout the United States also played a role in economic impact of AI patenting activity. A review by the USPTO of granted AI patenting activity for 25 years from 1976-2000 indicates that AI inventors and patent applicants were frequently located in technology hubs such as Silicon Valley, California and Boston, Massachusetts. The same study conducted between 2001-2018 indicates that AI patenting activity held steady in Silicon Valley and throughout the Northeast, as well as in other areas of the country including Florida, Arizona, Texas, Colorado, Illinois, Michigan, Washington, and Oregon.
The results of this study indicate a profound explosion in the utilization and adoption of AI based technologies throughout the United States. We can expect to continue to see continued patenting and innovation in this space, driven by the ability of innovators and entrepreneurs to continue to adopt and incorporate AI based innovations into offered products and services. As patent activity increases, we can expect the need for patent claim breadth and coverage to be amended to get around an increasing quantity of patent applications and publications in this technical space. Such needs make working with effective patent counsel to create a patent strategy to robustly protect and defend innovation in this area imperative.
 NIST (2019), 7-8. In a leading textbook, Russell and Norvig (2016) define AI broadly as the development of machines capable of under-taking human activities in four areas: thinking humanly, acting humanly, thinking rationally, and acting rationally.