By Charlie Damschen | Guest Column
Nearly everyone is aware of artificial intelligence (AI) at some level, even if it’s only through one of Hollywood’s various doomsday scenarios (“The Terminator,” “The Matrix,” etc.).
While the prospect of a Skynet-led takeover of the Earth doesn’t exactly inspire confidence in the future of technology, the reality is that the potential to increase global productivity and generate tremendous wealth through the use of AI is extremely high. A recent analysis by Gartner projects that AI will directly contribute to a $2.9 trillion increase in business value and 6.2 billion hours of worker productivity globally within the coming year alone.
As with any valuable asset, businesses are looking for ways to protect their investments – in this case, to prevent copycats from using their valuable AI assets without investing resources to develop their own.
Enter the U.S. Patent & Trademark Office. Not only have the number of patent applications that specifically mention AI greatly increased in the last year, but the number of granted applications has also greatly increased. That means the U.S. Patent Office is issuing more and more patents directed to various AI-related inventions. In fact, the number of issued patents on AI-related inventions nearly doubled in 2019 compared to 2018.
As expected, global tech giants have become the top three recipients of AI-related patents. IBM comes in at number one, followed by Microsoft and Amazon. Somewhat surprisingly, number four on the list is Capital One, while Japan-based Fanuc Corp., a global producer of robotics and CNC systems, received more AI-related patents in 2019 than Cisco, Facebook, Google or Apple.
Clearly the financial industry has an interest in proprietary rights to AI-related technology. Furthermore, it is safe to assume that companies are trying to protect at least some number of AI inventions as trade secrets, rather than by obtaining a patent. Accordingly, the total amount of resources directed to developing AI-related assets is growing rapidly across a broad spectrum of industries.
These AI-related inventions logically lead to the question, “What if the AI itself invents something?” It’s an intriguing question without a firm legal answer as of yet.
At the end of 2019, the U.S. Patent Office put out a request for public comment on the subject of patenting AI-related inventions. The request included 13 different questions and directly addressed whether a work produced by an AI algorithm or process, without the involvement of a natural person, should be protectable under either copyright law or patent law. A related issue included in the office’s request for public comment is how to treat creations (either through copyright or patent law) that are a result of a collaboration between AI and a natural person.
Just last month, the European Patent Office (EPO) rejected two patent applications that listed AI as the inventor. The EPO reasoned that the framework of the European patent system indicates that the inventor must be a natural person.
As more resources are used to develop more sophisticated AI, one can expect an increase in the number of patent applications (both domestic and internationally) directed to that AI. So, it appears that in the near-term this issue is going to be an important one for determining patent rights and patent validity related to AI inventions.
To that end, business leaders should be encouraged that the U.S. Patent Office has requested input from the public regarding its treatment of AI inventions. Hopefully the responses reflect multiple stakeholders in various roles, and not just the views of big tech and financial companies. The U.S. Patent Office’s implementation of polices in light of the responses it received is still unknown, but the mere fact that it solicited input from the public will hopefully lead to policies that benefit a broader spectrum of stakeholders.
Charlie Damschen is a partner at Hamilton IP Law, with offices in Coralville and Davenport.