AI and Patents – "the person skilled in the art"
Or AI and the State of the Art
02 August 2019
"Artificial intelligence" is a broad term used to describe a range of computer and software functionality. There is no single definition of AI. It is often misused to describe automated processes, but most commonly it is used to describe different levels of machine learning.
In previous articles we have looked at the UK copyright implications of AI. In this series of posts we look at some of the implications for patent law, and in particular how the outputs from autonomous forms of AI may be affected.
The "person skilled in the art"
Before we look at the possible impact of AI on the definition of the person skilled in the art, it is worth briefly outlining the concept. It is crucial to assessing the patentability of an invention. An invention is only capable of being patented if it involves an "inventive step". Put another way, if the invention was obvious to the person skilled in the art it is not be capable of protection.
The definition of the person skilled in the art will vary on a case-by-case basis. Background knowledge and technical ability will vary from industry to industry and invention to invention. The "person" can in fact be a team of people with different skills. In each case the person is deemed to be a skilled technician, but with a lack of imagination or inventiveness. The person is also deemed to be aware of the "common general knowledge", such knowledge being that known to ordinarily skilled or qualified individuals in the relevant technology field.
As has been well documented, one of the major benefits of the use of AI at present is the ability to process huge amounts of data in a relatively short period of time. What if this vast processing power is deemed to be accessible to the person skilled in the art? Could that have the effect of changing the current concept of what is inventive over prior art, or elevating the ability of the unimaginative and uninventive technician? Does it shift the focus on to whether the person skilled in the art would have used an AI system in the same way to try to resolve a problem?
The way obviousness is approached is to first identify the person skilled in the art and the relevant common general knowledge. Then the inventive concept claimed must be identified, and the differences between the state of the art and the inventive concept. Finally, the test to be applied is whether without any knowledge of the alleged invention, are the differences between the state of the art and the inventive concept steps obvious to the person skilled in the art or do they require any degree of "invention"? If the inventive step can be reduced to the application of an AI system to a particular problem, arguably it may make it harder to show that the invention involved an inventive step. This is likely to be a bigger issue in fields where "invention" arises from the assimilation of large bodies of data or fields ordinarily requiring considerable investment in experimentation, such as the pharmaceutical industry. The value of patents to those fields demonstrates the enormity of the issues at stake.
At present, AI systems are seen as a sophisticated tool aiding invention. This has been cited as one of the justifications for there being no need to reassess whether a non-human actor could be considered an inventor, discussed further in this article. As the use of AI systems becomes more prevalent, and as these systems become more advanced, it will become harder to show that the human intervention in the creation of a new invention did in fact require any degree of invention. The courts may have to recalibrate the way they think about the person skilled in the art, or what is in fact "invention", to take that into account, since the alterative of recognising the AI system as inventor is far more drastic and controversial.