AI patents still possible in the UK despite Court of Appeal ruling

Harry Muttock, a patents expert at Pinsent Masons, was commenting after the Court of Appeal in London overturned an earlier ruling of the High Court, which had considered that a trained ‘artificial neural network’ (ANN) – a form of AI system – was patentable.

Muttock said that while the Court of Appeal’s ruling was not helpful for software developers seeking to patent their inventions in the UK, ultimately, it merely restored the UK position to where it was prior to the High Court’s judgment in this case.

“There is a perception within industry that patents for software inventions are harder to obtain in the UK than they in fact are,” Muttock said.

“The High Court’s ruling had suggested that trained software would be easier to patent than other types of software, but while the Court of Appeal has now clarified that that is not the case, it does not mean it is any harder for AI developers to patent their inventions than it is for other types of software. While there are hurdles to clear, as this new ruling highlights, software patents are possible, are an option, and can be commercially advantageous,” he said.

The case before the Court of Appeal concerned a long-standing dispute between Emotional Perception AI Ltd (EPAI) and the UK’s Intellectual Property Office (IPO).

EPAI has developed a system that it claims makes music and other media recommendations to users in an improved way, based on how the users have been categorised by trained artificial neural networks.

The High Court described an ANN as an aspect of AI system and as being “a black box which is capable of being trained as how to process an input, learning by that training process, holding that learning within itself and then processing that input in a way derived from that training and learning”.

The case originally came before the High Court after a hearing officer at the IPO rejected EPAI’s patent application for its system. The hearing officer considered that the EPAI system as a whole was a computer program and that its provision of “semantically similar file recommendations” was “not technical in nature”, meaning he considered it was excluded from patentability under UK patent law.

High Court judge Sir Anthony Mann disagreed, however. He considered that the patent claimed by EPAI was not for a computer program at all, and therefore that the associated exclusion to patentability under the Patents Act did not apply.

Sir Anthony Mann said it was appropriate to distinguish between the two different types of ANN that operate within EPAI’s system. It is, he considered, only “the program which achieves, or initiates, the training” – which he characterised as the ‘hardware ANN’ – that constitutes a computer program. Even so, he found that the hardware ANN provided a ‘technical contribution’ that lifted it out of the Patents Act’s exclusion on patentability for computer programs. This, he said, was because the system produces an output – a file – that “would not otherwise be selected”.

In reaching that view, Sir Anthony suggested that a computer-implemented method for selecting non-technical criteria may fulfil the requirement of a technical effect if the selection deriving from the application of the criteria is technical in nature. 

The High Court judgment prompted the IPO to update guidance on examining patent applications relating to AI inventions, but Muttock said the guidance will now need to be revised again after the Court of Appeal overturned Sir Anthony’s ruling.

The Court of Appeal first held that ANNs constitute computer programs. It considered the weights and biases that apply to inform how an ANN operates in reaching that assessment.

“However it is implemented, such a machine is clearly a computer – it is a machine for processing information,” Lord Justice Birss said in his leading judgment for the court. “Focussing on the weights of an ANN, in my judgment irrespective of the manner in which an ANN is implemented (hardware or software), … these weights are a computer program. They are a set of instructions for a computer to do something.”

“For a given machine, a different set of weights will cause the machine to process information in a different way. The fact the set does not take the form of a logical series of 'if-then' type statements is irrelevant. The weights for a given artificial neuron are what cause the neuron, if the inputs are of a given type, to then produce an output of a given type. Aggregated up to the ANN as a whole, these weights work that way in parallel with one another to a significant extent and not just in a logical series, but that is not a relevant distinction. The set of weights as a whole instruct the machine to process information it is presented with in a particular way,” he said.

While the court held that ANNs fall into the exclusion on patentability applicable to computer programs, it acknowledged that some computer programs have been, and can be, lifted out of the exclusion if they can be shown to make a technical contribution. In this case, the Court of Appeal determined that no such technical contribution applies in the context of EPAI’s ANNs.

“It is true that … the system has gone about its analysis and selection in a technical way but that is because it is an ANN, i.e. a computer,” Lord Justice Birss said. “The fact the computer is using properties it can measure to make this semantic recommendation makes no difference. I think the flaw is that this approach imports the undoubtedly technical nature of computer systems (including ANNs) into the analysis… It is the semantic similarity of the files here which gives rise to their recommendation but that is not a technical matter at all.”

“Putting it another way the similarity or difference between the two files is semantic in nature and not technical… The fact that in the present case there is what one might call an external transfer of data (the file recommendation) does not help ... What matters is the correct characterisation of the data being transferred and that brings the issue back to the aesthetic and therefore non-technical quality of this aspect of the contribution,” the judge said.

Muttock said the Court of Appeal’s ruling is relevant given the wider industry context.

“Businesses are developing new adaptive and iterative software systems and in tandem assessing the UK’s position on the patentability of those systems to better understand how attractive and rewarding the UK is for innovators and developers in this area,” he said.

“This ruling shows that trained ANNs are to be treated no differently to any other computer program for the purposes of assessing their patentability in the UK. The Court of Appeal’s judgment can be read as extending to other forms of trained software too, which has implications for developers of other ’AI’ systems. The IPO will likely seek to revise its guidance on examining patent applications again in light of this ruling, with it likely to revert back to something closer to the wording it had prior to the High Court’s judgment in this case,” he said.