Stephen Thaler and Ryan Abbott plan to bring a light beacon, a beverage container, and a machine called Dabus into court, along with a simple question: Does an inventor need to be human?
Depending on how they respond, a panel of judges on the U.S. Court of Appeals for the Federal Circuit could open the door to another significant question: What is “obvious” to a machine?
A basic tenet of U.S. law is that patents aren’t awarded for inventions that are obvious. The standard of obviousness in patent law is measured against a hypothetical person of ordinary skill in the art.
Putting artificial intelligence, with its potential for near omnipotent capabilities, on equal footing as human inventors could have a significant impact on patent law’s obviousness standard, attorneys and patent professionals say.
“You naturally gravitate to the question of: ‘If Watson is now on the playing field as a person of ordinary skill in the art, isn’t it going to become true that everything is obvious?’” said Chris Mammen, an attorney at Womble Bond Dickinson.
The Dabus appeal is viewed by many as a longshot. Arguments for recognizing artificial intelligence as an inventor have been rejected by most patent offices and courts that have addressed the question. At the district court, a U.S. judge relied on Federal Circuit statements that “only natural persons can be inventors.”
But even without recognizing artificial intelligence as an inventor, the technology could require the obviousness standard be recalibrated, some say, as new technologies help researchers and raise the capabilities of humans of ordinary skill.
“It should be making the obviousness standard a harder bar to overcome,”said JiNan Glasgow George, CEO of patent analytics firm Patent Forecast.
AI as Inventor
Dabus is an artificial intelligence creativity machine that “invented” a fractal container and a “device for attracting enhanced attention.” Dabus was built by Thaler, a computer scientist.
Abbott, a law professor at the University of Surrey in England, and a team of attorneys have filed patent applications in numerous countries listing Dabus as the inventor.
A judge in the Eastern District of Virginia last month upheld U.S. patent office rejections of the applications. Thaler and Abbott have appealed the ruling to the Federal Circuit. Their written arguments are due in late November.
Recognizing artificial intelligence as an inventor could have broad legal ramifications. With respect to the obviousness standard, the American Intellectual Property Law Association has said the impact would likely be “profound.”
“What seems like a magical invention to a human could be rather obvious to an AI machine because it has the capability to crunch through a bunch of numbers in a very fast period of time and come up with an answer to a problem in minutes that would take a human being a lifetime,” said Marc Richards of Crowell & Moring.
Abbott said in a 2017 academic paper the standard for patentability will be too lenient if it fails to evolve. When, or if, artificial general intelligence is developed, “inventive machines” should become the skilled person in all areas, he argued.
Others are leery, noting the obviousness standard has always been capped at an expectation of a human of ordinary creativity.
“Once you take that cap off, or start thinking about taking that cap off, that’s where you start to go down the path of, ‘Which standard are we judging?’” Mammen said.
AI as Tool
Drug companies use artificial intelligence-based tools to identify promising drug designs and to analyze molecular characteristics of patients, which can help scientists select the correct treatment. Machine learning and pattern recognition software can also help researchers identify problems and solutions.
By making certain tasks easier, it raises the question of whether artificial intelligence is increasing what someone of ordinary skill in the art is capable of, and whether it should be accounted for in the obviousness standard. Parallels have been drawn to the internet.
“Defining what is skill in the art will be evolving, as it was decades ago,” said Kate Gaudry, a lawyer with Kilpatrick Townsend & Stockton. “Now it’s kind of assumed a person had access to a computer and the internet.”
But artificial intelligence isn’t ubiquitous yet. While utilized in areas like drug development or cybersecurity, AI aides aren’t prevalent in things like automobile or surgical device design, attorneys say.
Taking the capabilities of artificial intelligence into account could make it “impossible for everyday inventors without access to AI to make a patentable contribution to their respective, far-ranging fields,” the SUNY Research Foundation has warned.
The International Association for the Protection of Intellectual Property suggests artificial intelligence’s capabilities only be considered to the extent that artificial intelligence software is readily available to someone of ordinary skill in the art, and that person is able to use the software.
“At some point, it should overtake all,” George said. “But is it today? I think probably not yet.”
— With assistance from Susan Decker, Bloomberg News.