On 22 September 2022, the IPO published its guidance on the examination of patent applications for inventions relating to artificial intelligence (AI). This follows the UK Government’s consultation in relation to the interaction of artificial intelligence with patent law and copyright law.
The guidelines define AI as “technologies with the ability to perform tasks that would otherwise require human intelligence, such as visual perception, speech recognition, and language translation”.
As AI inventions typically rely upon mathematical models and algorithms for their implementation, this guidance emphasises the importance of examining applications in light of the exclusions for mathematical methods and computer programs under UK patent law.
The guidance confirms that an AI invention will avoid exclusion if it makes a technical contribution to the state of the art beyond that consisting purely of excluded subject matter. When assessing this, the examiner will consider five “sign-posts” (based on the reformulated test in HTC v Apple  EWCA 451):
- “whether the claimed technical effect has a technical effect on a process which is carried on outside the computer;
- whether the claimed technical effect operates at the level of the architecture of the computer; that is to say whether the effect is produced irrespective of the data being processed or the applications being run;
- whether the claimed technical effect results in the computer being made to operate in a new way;
- whether the program makes the computer a better computer in the sense of running more efficiently and effectively as a computer;
- whether the perceived problem is overcome by the claimed invention as opposed to merely being circumvented.”
Practical example: “AI genome”
There is currently only one case in which the UK court has considered the patentability of an AI invention in respect of excluded subject matter.
In Reaux-Savonte v Comptroller-General of Patents  EWHC 78 (Ch), the Judge held that an “AI genome” was excluded from patentability. In reaching this decision, the Judge considered that the invention, which was understood to be an arrangement of computer code which facilitated its evolution over time, allowing it to adapt and change in the same way that biological code evolves, was excluded as it failed to meet signposts iii) or v).
Illustrative examples: life sciences
Although there is very little case law in this area, the guidance is accompanied by a number of helpful illustrative examples.
One of these considers an application for an AI invention, which includes the method of training a neural network with heart imaging data sets and the use of the trained neural network to measure the percentage of blood leaving the heart.
In this example, the invention would not be excluded from patentability on the basis of excluded subject matter, as the contribution related to an improved measurement of the percentage of blood ejected from the heart, i.e. that it showed a technical effect upon a process lying outside the computer that implemented the invention.
Despite the limited case law in respect of AI inventions, the guidance and the incorporation of the five sign-posts provide useful clarity as to the criteria an AI invention must meet to avoid exclusion from patentability.
The illustrative examples are also a handy resource in understanding how the guidelines may be applied in practice, and where the threshold lies for demonstrating a technical contribution beyond that which consists of purely excluded subject matter.