The UK Intellectual Property Office (UKIPO) has issued new guidance on the search and examination of UK patent applications for computer-implemented inventions, including inventions involving artificial intelligence: Search and Examination of UK Patent Applications under the Patents Act 1977 (as amended).

The guidance was issued in response to the UK Supreme Court’s judgment in the ongoing Emotional Perception case, which has now returned to the examination stage: Emotional Perception AI Limited v Comptroller General of Patents, Designs and Trade Marks.

The guidance confirms that, when assessing the patentability of a claim, UK examiners should follow the UK Supreme Court’s interpretation of Section 1 of the UK Patents Act in the Emotional Perception judgment. In particular, examiners should no longer apply the so-called “Aerotel test” to determine whether the invention claimed falls wholly within excluded subject matter.

More specifically, UK examiners should follow a three-step approach consistent with the EPO’s approach to assessing patentability, comprising a first hurdle, an intermediate step and a second hurdle. For computer-implemented inventions, the first hurdle can readily be overcome by including “any hardware” in the patent claim—for example, by specifying that the method steps are performed by a computer.

The intermediate step requires identifying which features of the claimed invention contribute to its overall technical character. This requires each feature of the claim to be analysed individually. Even if a feature is not technical when considered in isolation, it may nevertheless contribute to, or interact with, the technical subject matter of the claim.

The second hurdle involves assessing whether the claim is novel and involves an inventive step. Although the guidance indicates that the first hurdle and the intermediate step are likely to be applied in a manner consistent with the EPO’s application of the corresponding steps, the same is not true of the second hurdle. Specifically, in accordance with the Supreme Court’s judgment, the guidance indicates that UK examiners should continue to follow the established Pozzoli approach when assessing inventive step, rather than adopting the EPO’s problem-and-solution approach.

Although this difference between the UKIPO’s and EPO’s approaches to assessing inventive step is not confined to computer-implemented inventions and applies to patent applications across all fields of subject matter, it may lead examiners and patent practitioners before the two offices to form different understandings and judgments when considering the “technical contribution” or “technical effect” of a claim.

Under the EPO’s problem-and-solution approach, an examiner first identifies the features that distinguish the claim from the closest prior art and considers the technical effects produced by those features. On that basis, the examiner formulates the objective technical problem to be solved. The examiner then considers, from the perspective of the person skilled in the art, whether that person, starting from the closest prior art and faced with the objective technical problem, would have modified the closest prior art in an obvious manner to arrive at a solution falling within the claimed subject matter.

Under the UKIPO’s Pozzoli approach, the inventive concept of the claim must first be identified. For a computer-implemented invention, only the technical features of the claim may be considered in the inventive-step assessment. In accordance with the guidance in paragraph 106 of the Emotional Perception judgment, the inventive concept may correspond to the technical character of the claim as a whole, as identified during the intermediate step. The cited “prior art” must then be compared with the inventive concept of the claim to determine whether any differences exist and, if so, what those differences are.

It follows that, for the purposes of inventive-step assessment, the “inventive concept” identified by the UKIPO differs fundamentally in nature from the “technical effect” identified by the EPO. This is because the EPO determines the “technical effect” by reference to the closest prior art, whereas the UKIPO identifies the “inventive concept” without reference to the prior art. This distinction is particularly noteworthy in the field of computer-implemented inventions, where the “technical character” of an invention is itself often contentious. Whether this distinction will ultimately produce different examination outcomes at the two offices remains to be seen.

Second, the EPO’s “technical effect” generally concerns what is achieved by the novel features of a claim—in other words, the benefit they produce—rather than the claim features themselves. By contrast, the UKIPO’s “inventive concept” is generally expressed as a combination of the claim features and their effects. For example, the inventive concept may lie in key features A, C and F, which produce benefit or effect X. The UKIPO’s usual understanding of a claim’s “inventive concept” therefore appears not to be entirely equivalent to the technical character of the claim as a whole. Instead, it tends to refer to one or more key technical features combined with their asserted beneficial technical effect. Indeed, to carry out the third step of the Pozzoli test—which requires a comparison between the prior art and the inventive concept—it is necessary to identify the claim features that contribute to the inventive concept, rather than merely their technical effects. This ensures that the comparison is made between features, instead of attempting to compare features with effects.

Overall, the new guidance represents a welcome step towards bringing the UKIPO’s approach closer to that of the EPO, whose approach often provides applicants with relatively predictable outcomes.