The buzz around artificial intelligence is generating an increasing number of opinions on how the proliferation of AI is going to dramatically change the Intellectual Property landscape. In particular, how businesses, intellectual property practitioners, public agencies and legislators are dealing with the dawn of the AI age. Let’s have a look.
In IFI Claims’ annual ranking of top patent recipients in the US for 2016, IBM sits in first place with a record-breaking 8,088 patents granted for the year, nearly 10 percent more than in 2015 . According to IBM, more than 2,700 of these patents are for inventions related to AI, cognitive computing, and cloud computing. IFI Claims is a patent services company that uses United States Patent and Trademark Office (USPTO) data to compile listings.
The boom of AI-related patent applications should mean a lot more work for the USPTO and other patent granting authorities worldwide in the coming years. This is hurdle number one: how to meet the exponential demand? Could it be through automation of the review and approval process for patent and other IP registration applications? In other words, through AI itself? Automation of patent searches and related operations are already out there.
Not only will patent offices have to expand to cope with the growing number of demands, but they may also require new ways of interpreting IP law to determine whether a particular AI-related invention, including a possible invention by artificial intelligence itself, is eligible for a patent, and who the inventor is. Courts and intellectual property lawyers are already dealing with these questions, as increasingly, AI is becoming a necessary tool for litigation. John Fernandez, chief innovation officer at Dentons, said recently in The New York Times: “Our industry is being disrupted, and we should do some of that ourselves…”
IP laws, like many of the laws in our society, were first drafted eons ago, well before things like the internet and the popularity of AI. Laws today don’t clearly address some of the questions that are puzzling businesses and the IP legal community, such as whether a non-human can qualify as an inventor, how will a company know that its artificial employee has created something worthy of a patent, and generally, what are the conditions for patentability of an AI-related invention?
Even though intellectual property laws are updated from time to time, the propagation of artificial intelligence may call for a more fundamental change, not only in definitions but also in the way that we characterize innovation and creativity at the heart of IP legislation. For example, can you automate the creative process? Arguably, once you do, it’s no longer creative. Another interesting question is what happens if someone uses AI for less than upstanding purposes? You can’t charge a machine or a program with a crime so what’s the recourse? These legal uncertainties are real and have to be addressed.
Because intellectual property legislation is meant to incentivize innovation by rewarding creators and protecting their investments, there may not be sufficient motivation for IP creation without IP protection. All of these issues will eventually force governments to act, and possibly reinvent IP laws to adapt them to our new reality. If they don’t act quickly enough, we could witness the most dramatic erosion of IP that the modern era has seen thus far.