UK Supreme Court Aligns with EPO “Any Hardware” Approach in Emotional Perception AI Patentability Decision

The UK Supreme Court today handed down its judgment in Emotional Perception AI Limited v Comptroller General of Patents, Designs and Trade Marks: Emotional Perception AI Limited (Appellant) v Comptroller General of Patents, Designs and Trade Marks (Respondent) - UK Supreme Court.

 

The original issue in this case related to the question of whether an artificial neural network (ANN) is a ‘program for a computer’ within the meaning of Article 52(2)(c) EPC (Article 52 – Patentable inventions). Rather unsurprisingly, the Supreme Court has answered this in the affirmative, meaning that patent claims directed to ANNs as such are excluded from patentability.

 

However, the more consequential part of today’s decision relates to another issue considered by the Supreme Court as part of this case. Namely, should the current guidance followed by the UK Intellectual Property Office and UK Courts for assessing whether claimed subject matter is an ‘invention’ under Articles 52(2) and 52(3) EPC continue to be followed? The answer is no.

 

The current guidance was established a couple of decades ago in Aerotel Ltd v Telco Holdings Ltd [2006] EWCA Civ 1371; [2007] RPC 7 (‘Aerotel’). At that time, it was considered that there was no settled approach in the EPO’s Boards of Appeal for making the above assessment in relation to Articles 52(2) and 52(3) EPC. The Enlarged Board of Appeal decision G1/19 [2021] EPOR 30 (‘G1/19’) a couple of years ago changed this. G1/19 endorsed the ‘any hardware’ approach used by the European Patent Office. It also, by implication, endorsed the criticisms of the Aerotel approach outlined in EPO Board of Appeal decision T 154/04 (‘Duns’). In particular, Duns rejected the Aerotel approach as being incompatible with the European Patent Convention.

 

Now that this issue seems to be settled at the EPO, then the UK Supreme Court feels obliged to follow. The Aerotel approach is therefore out, and the EPO ‘any hardware’ approach is in. So far, so simple. But questions arise as to the implications of this change, particularly around the assessment of inventive step, as acknowledged by the Supreme Court in today’s decision.

 

The EPO’s approach to assessing patentability follows the so-called ‘two hurdle’ approach. The first step of this approach considers whether the claimed subject matter is an ‘invention’ within the meaning of Articles 52(2) and 52(3) EPC. In practice, this is a relatively low hurdle to overcome because if the claim recites ‘any hardware’ then it is regarded as not reciting excluded subject matter. If the first step is passed, then the second step considers whether the claimed invention is inventive. The EPO uses its long-established problem-and-solution approach to perform this assessment.

 

For claims that involve a mixture of technical and non-technical features, such as many computer-implemented inventions, an intermediate step is performed before the second step of the two hurdle approach. This intermediate step identifies the technical effect of the claimed invention and, in particular, identifies which features of the claim contribute to the technical effect, noting that features that are in and of themselves non-technical can still contribute to the overall technical effect achieved by the claimed invention. For the second step of the two-hurdle approach, only those claim features that contribute to the technical effect can be taken into account in the assessment of inventive step. This is the ‘COMVIK approach’ from EPO Board of Appeal decision T 641/00 (T 0641/00 (Two identities/COMVIK) 26-09-2002 | epo.org).

 

The UKIPO and UK Courts use a different approach for assessing inventive step, namely, the so-called ‘Pozzoli’ approach (Pozzoli SpA v BDMO SA [2007] EWCA Civ 588; [2007] FSR 37). The question is, for computer-implemented inventions, whether the Pozzoli approach is compatible with today’s judgment that the UKIPO and UK Courts need to follow the ‘any hardware’ approach for assessing excluded subject matter before moving on to assess inventive step.

 

The Supreme Court considers that there is no need for the UK Courts to change their approach to the analysis of what amounts to an inventive step, although it notes that some modification many nonetheless be needed. However, the Supreme Court does consider that an ‘intermediate step’ as n the COMVIK approach is needed to ensure that claim features that do not contribute to the technical character of the claimed invention are excluded from the assessment of inventive step in the second step of the two hurdle approach. The Supreme Court considers that the UK Courts are free to determine how to implement the intermediate and second steps of the two hurdle approach, such that the Pozzoli approach could be used to implement the second step, rather than using the EPO’s problem-and-solution approach.

 

The UKIPO and UK Courts will need to decide how to implement this intermediate step. And decide how the outcome of this step can be reconciled with using the Pozzoli approach to then assess inventive step. It is suggested in today’s judgment that the ‘inventive concept’ identified as part of the Pozzoli approach could be used as a method of identifying the technical character of the invention as part of the intermediate step. But whereas in the problem-and-solution approach the identified technical effect is based only on claim features that are novel over the identified closest prior art, in Pozzoli the ‘inventive concept’ is identified without regard to the prior art. An open question is therefore how to perform the intermediate step – as now required by the UKIPO and UK Courts – in a manner that: 1/ does not also require using the problem-and-solution approach to perform the second step; and that allows for the Pozzoli approach to continue to be used by the UKIPO and UK Courts to assess inventive step as part of the second step of the two-hurdle approach.