On 21 November 2023, Mann J handed down his judgment on Emotional Perception AI Ltd v Comptroller-General of Patents, Designs, and Trade Marks [2023] EWHC 2948 (Ch)i, relating to the patentability of an invention relating to artificial intelligence.
At first glance, the judgment may appear to pave the way for a new standard of examination for computer-implemented inventions at the UKIPO, but caution should be exercised.
The invention
Broadly speaking, the invention is related to a process by which, based on an input media file such as a music file or a video file, an artificial neural network is used to identify another media file which is semantically similar to the input media file. To use the example of a music file, a semantically similar music file might be one which evokes a similar emotional response, or which conveys a similar message, rather than one which has a similar rhythm, melody, or tempo.
The claim which was considered essentially included two separate processes: a process by which an artificial neural network is trained, and a process by which the trained artificial neural network is used.
Background to the judgment
This case was an appeal of a UKIPO Hearing Officer Decision (BL O/542/22ii). In that Decision, the Hearing Officer rejected the application as being directed towards excluded subject-matter under section 1(2) as a computer program as such.
This conclusion was based on the notion that the provision of a better (i.e. more semantically similar) recommendation is an effect which is “of a subjective and cognitive nature and does not suggest that there is any technical effect over and above the running of a program on a computer”.
Briefly, in the High Court, two questions were considered:
- Is the invention a claim to a computer program at all?
- If yes, is there a technical contribution beyond that?
The judgment
The answers provided by Mann J were at odds with the current approach of the UKIPO to examination of computer implemented inventions, and therefore likely to be surprising for with experience prosecuting patent applications for computer-implemented inventions in the UK.
Mann J considered that the process by which the artificial neural network is trained is a program for a computer as such. However, he considered that, even when implemented in software, the trained artificial neural network itself is not a program for a computer, and therefore that the claim as a whole could not fall under the exclusion.
Having concluded that the exclusion did not bite, Mann J stated that it was not necessary to consider whether the invention makes a technical contribution, but did so anyway.
Surprisingly once again (perhaps even more so), he found that the claimed invention did make a technical contribution.
The reasoning is, in my view, so markedly different from what we are used to seeing from UKIPO examiners, that it is worth reproducing in full:
The correct view of what happened, for these purposes, is that a file has been identified, and then moved, because it fulfilled certain criteria.True it is that those criteria are not technical criteria in the sense that they can be described in purely technical terms, but they are criteria nonetheless, and the ANN has certainly gone about its analysis and selection in a technical way. It is not just any old file; it is a file identified as being semantically similar by the application of technical criteria which the system has worked out for itself. So the output is of a file that would not otherwise be selected. That seems to me to be a technical effect outside the computer for these purposes, and when coupled with the purpose and method of selection it fulfils the requirement of technical effect in order to escape the exclusion. I do not see why the possible subjective effect within a user's own non-artificial neural network should disqualify it for these purposes. To adapt the wording of Floyd J in Protecting Kids, the invention is not just one depending on the effect of the computerised process on the user. There is more than that. There is a produced file with (it is said) certain attributes. The file produced then goes on to have an effect on the user (if the thing works at all) but one cannot ignore the fact that a technical thing is actually produced. It would not matter if the user never listened to the file. The file, with its similarity characteristics, is still produced via the system which has set up the identification system and then implemented it.
Based on this reasoning, the appeal was allowed, and the Decision of the Hearing Officer was overturned.
What does this mean for AI inventions at the UKIPO?
This is certainly a positive development for the patentability of AI inventions, and computer-implemented inventions more generally. However, care should be taken, at least for now.
The approach adopted by the High Court here appears to diverge somewhat with the EPO’s beloved “problem and solution” approach for the assessment of inventive step. Based on current EPO guidelines, it is hard to imagine an that EPO examiner would acknowledge a technical purpose in identifying semantically similar media files, regardless of whether a file (to use Mann J’s wording: “not just any old file”) can be considered a “technical” object. However, the EPO guidelines are amended regularly – and was recently amended to include clear and relatively generous provisions for applications of AI that can be considered technical according to the EPO’s approach. On the other hand, this conclusion does share similarities with approach taken in the US, where if the machine-learning is complicated enough that it could not be done in the human mind, it may not fall foul of the §101 objection to abstract subject-matter.
We do not yet know how the UKIPO will respond to the judgment. It is possible that they will appeal to the Court of Appeal, in which case the judgment may be overturned. If so, then the status quo would remain. On the other hand, if the judgment were confirmed by the Court of Appeal, this could indeed mark the beginning of a more AI-friendly era at the UKIPO, albeit one which might lead to different results from their European counterparts.
Something which, for procedural reasons, was not considered in this case was whether the artificial network would have been excluded as a mathematical method as such, which is the objection more frequently encountered. It would have been interesting to see how this question would have been answered. Even so, this appears to be somewhat of a moot point, given Mann J’s positive finding in respect of a technical contribution.
All in all, we can be cautiously excited, at least for the time being, that this case law will greatly improve the chances of success for patent applications for computer-implemented methods at the UKIPO, particularly those directed towards developments or applications of AI.
References:
- https://www.bailii.org/ew/cases/EWHC/Ch/2023/2948.html
- https://www.ipo.gov.uk/p-challenge-decision-results/o54222.pdf
