Patentability of AI and Software: UK Supreme Court’s Emotional Perception Judgment Explained
The UK Supreme Court (UKSC) has reformed the way that the patentability of an invention is assessed. The updated approach should be applied whenever a determination is being made of whether an invention is excluded from patent protection. The Emotional Perception judgement has fundamentally changed the way that computer implemented inventions (CII) are examined, especially technology that relates to artificial intelligence (AI) and artificial neural networks (ANN).
Key Takeaways
UK courts are expected to be aligned with European law when assessing the patentability of a CII invention, including AI inventions. This consistent approach should improve legal certainty when pursuing a UK patent, regardless of whether it is filed at the UK Intellectual Property Office (UKIPO) or at the European Patent Office (EPO). A patent application should explain the technical effects of the innovation that is described.
Is this judgement welcome?
Yes. It is important for the law to keep up with technological developments. The consequences of this judgement will be of relevance to anybody who is interested in pursuing patent protection for CII inventions including AI inventions.
The UKSC has answered longstanding questions, while also opening up some new questions. UKIPO examiners, in correspondence with UK patent attorneys, are leading the way in interpreting the Emotional Perception judgement. They will be closely followed by the UK’s specialist patent courts. The emerging legal landscape will shape the way that patents and patent applications are to be examined from now on. For some of the questions that have been raised by this judgement, it will likely take many years for UK case law to become settled. In the meantime, below we offer some tentative answers. The [paragraph numbers] below identify the parts of the Emotional Perception judgement that we are relying on for our review.
What is the relevant statute that is being considered by the UKSC?
- Patentability is defined by Section 1 of the UK Patents Act 1977 (PA) [14].
- This is consistent with Article 52 of the European Patent Convention [14].
- Section 1 PA and Article 52 EPC are framed so as to have, as nearly as practicable, the same effects [13].
What is the significance of this judgement?
For the past 20 years, when determining what is a patentable invention, UK courts and the UKIPO have followed the test established by Aerotel v Telco Holdings [13, 29]. With this UKSC judgement, the four-step Aerotel test has been rejected [56-67]. Instead, UK law should apply decision G1/19 of the Enlarged Board of Appeal (EBA) of the European Patent Office (EPO) [113].
The Emotional Perception decision is binding on UK courts, as well as the UKIPO. This decision will be applied whenever the patentability is being considered for:
- a UK patent application,
- a UK patent that has been granted by the UKIPO, or
- a UK patent that has been granted by the EPO.
UK courts are not bound by EPO decisions. However, they would typically be expected to follow such decisions [44]. This is so even if the change might be disruptive [48-52]. The rejection of Aerotel has definitely disrupted the way in which excluded subject matter and inventive step are assessed in the UK.
What is the test under G1/19 of the Enlarged Board of Appeal (EBA) of the European Patent Office (EPO)?
When it comes to assessing patentability, in G1/19 (Pedestrian simulation), the EPO’s EBA approved a combination of:
- the “any hardware” approach, as set out in Duns Licensing Associates (Decision T 154/04), and
- the “two-hurdle" approach, for which the principles were established by the EPO’s Board of Appeal in Comvik (Decision T 641/00).
G1/19 endorsed the criticisms of Aerotel that were made in Duns Licensing [40, 56]. The “two-hurdle approach” is set out as follows [43]:
- Hurdle (i): Assess whether the claim amounts to an “invention”.
- Hurdle (ii): Assess whether the claim satisfies the conditions of novelty, inventive step and industrial application.
Patentability requires both of these requirements to be satisfied. Rather confusingly, G1/19 introduces the doctrine that the two-hurdle approach actually entails three steps [43]. Between hurdles (i) and (ii) above, there is an “intermediate step” of establishing whether a feature contributes to the “technical character” of the invention [43].
What does it mean under UK law for an invention to have “technical character”?
UK case law is not settled on the answer to this question. The fact that there is no precise test under UK law for determining whether a computer program is excluded from patent protection gives some wiggle-room, for arguing one way or another whether a claimed invention does or does not contribute technical character. When preparing to pursue patent protection of a CII, it is worthwhile taking the time to identify arguments that can be used to head off an excluded subject matter objection, or to respond if an excluded subject matter objection is raised.
To help assess whether a computer program makes a technical contribution, UK examiners typically apply five so-called AT&T signposts, as have been restated in HTC v Apple. The signposts are as follows:
- Signpost 1: Whether the claimed technical effect has a technical effect on a process which is carried on outside the computer (from Vicom T 0208/84).
- Signpost 2: 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 (from IBM T 0006/83 and IBM T 0115/85).
- Signpost 3: Whether the claimed technical effect results in the computer being made to operate in a new way (from Gale’s Application).
- Signpost 4: Whether the program makes the computer a better computer in the sense of running more efficiently and effectively as a computer (from Symbian; as reworded in HTC v Apple).
- Signpost 5: Whether the perceived problem is overcome by the claimed invention as opposed to merely being circumvented (from Hitachi T 0258/03).
When drafting a patent application for a CII, it is valuable to include arguments to demonstrate that the subject matter is “technical”. Regardless of whether “technical effect” is assessed for the purpose of determining excluded subject matter, or whether it is useful for demonstrating inventive step, it is worthwhile to include arguments that you can rely upon in your patent application as filed.
What does it mean under EPO law for an invention to have “technical character”?
Features that are “technical” are taken into consideration when deciding whether the claimed invention is inventive. Here, G1/19 is implementing the “problem-solution” approach, as was set out by Comvik (Decision T 641/00). This is the well-established test used by the EPO for assessing whether an invention satisfies the requirements of inventive step.
The “problem-solution” approach involves identifying, by reference to the closest prior art in the technical field of the invention, a technical problem which the invention claims to solve, and then assessing whether the technical feature(s) which form the claimed solution could be derived by the skilled person in that field in an obvious manner from the state of the art [32-33].
However, this does introduce a potential conflict between UK law and EPO law: UK case law does not apply the “problem-solution” approach for the assessment of inventive step. Instead, UK law applies the long-settled Pozzoli test, as was approved by the UKSC with the Actavis v ICOS decision [53], and reaffirmed by the UKSC with the present Emotional Perception decision [65].
The Comptroller General criticised any attempt to try to combine the approach in G1/19 with the settled English approach to inventive step as creating a Frankenstein monster [53]. This position is not accepted by the UKSC [67]. The Emotional Perception decision explains that neither Comvik, Duns Licensing or G1/19 suggest that the “problem and solution” approach of the EPO is the only way of assessing inventive step, or the only way of factoring the exclusions in Article 52(2) EPC into the overall appraisal of patentability [53-54].
A question that remains outstanding for another court to decide is whether the UK test of inventive step needs to be reformed [67, 116]. If this does happen, the EPO’s well-established “problem-solution” approach would be a strong candidate for replacing the Pozzoli test. Whichever of these tests for inventive step prevails, UK patent attorneys are well-versed in applying the different tests that have been formulated for assessing inventive step.
Issue 1: Should the Aerotel guidance no longer be followed [25-67]?
The UKSC answers this question by emphatically rejecting the Aerotel approach, and instead adopting the G1/19 approach [56, 64]. This decision taken by the UKSC is consistent with the position that was advocated by the Chartered Institute of Patent Attorneys (CIPA), which intervened in this case.
Issue 2: Is an ANN (or does it contain) a “program for a computer” [68-96]?
The UKSC concluded that an ANN is a program for a computer [96]. If an ANN with weights is considered to be a computer program, then the computer program exclusion could be broadened to include hardware ANNs. This concern is diminished by the UKSC recognising that it is common ground that the EBA in G1/19 approved the “any hardware” approach [31, 38]. This indicates that consumer electronics that implement ANNs would not be excluded from patentability.
The UKSC agrees with CIPA that the description of a computer as “a machine which processes information” is too broad [76]. What differentiates devices such as ovens and smoke detectors is that they are not used to perform computations.
The UKSC describes the machine on which an ANN is implemented as a computer, regardless of the technology that is used to achieve that functionality [86]. The ANN constitutes a set of instructions to manipulate data in a particular way so as produce a desired result, meaning that an ANN is considered to be a program for a computer [87].
Interestingly, the UKSC confirms that the term “computer” is broad enough to cover “quantum computers” [77]. This is consistent with the guidance that is already provided by the UKIPO (see Section 1.46 of the Manual of Patent Practice).
Issue 3: Is the entire subject matter of the claims excluded [97-99]?
The conclusion on issue 2 that an ANN is a program for a computer [96] makes it necessary to consider whether the entire subject matter of the claims falls within the scope of the exclusion of “programs for computers … as such” [97]. The rather odd phrase “as such” in the statute does a lot of the heavy lifting.
An assessment needs to be done of whether the claim includes any technical features and so qualifies as an “invention” within the meaning of Article 52(1) EPC [97]. This is hurdle (i) above of the “two-hurdle” approach [43].
The UKSC concludes that the UKIPO applied the Aerotel approach, rather than the EPO’s “any hardware approach” [98]. The UKSC explains that it could stop there, and remit the case back to the UKIPO [99]. Rather helpfully, the UKSC goes on to provide guidance on the way that the intermediate step between hurdles (i) and (ii) is to be assessed [100-111].
How should the intermediate step of G1/19 be followed [100-111]?
G1/19 approves the approach that was taken in Duns Licensing (T 154/04) [101], recognising that a claim may legitimately consist of a mix of technical and non-technical features.
Non-technical features “as such” do not provide a contribution to the prior art and are therefore to be ignored when assessing novelty and inventive step at hurdle (ii) of the “two-hurdle approach” [39, 101-103]. The intermediate step serves as a filter for features contributing to a technical solution of a technical problem, in view of the closest prior art. Only those distinguishing features can contribute to inventive step [43, 104].
The UKSC has left it open for the UKIPO and the UK courts to adopt any appropriate method of identifying the technical character of the invention [106]. A potential way of simplifying UK law would be to harmonise further with the EPO, by adopting the “problem-solution” approach for the assessment of inventive step. Whether this happens remains to be seen.
How should G1/19 be applied to the Emotional Perception case [112-118]?
The UKSC recognises that Emotional Perception’s patent application has not been examined by applying G1/19. This case has therefore been remitted back to the UKIPO for reconsideration by the Hearing Officer. Quite rightly, the UKSC has gone as far as is necessary to establish the relevant law that should be applied, and no further, affording the applicant the right to be heard on how the approach set out in G1/19 is relevant to the patentability of their innovation.
How harmonised does this leave UK law and EPO law?
By approving G1/19 and rejecting Aerotel, the Emotional Perception judgement harmonises the approach to patentability under UK law and EPO law. In fact, the UKSC goes further than that, paving the way for increased harmonisation in the way that excluded subject matter and inventive step should be assessed. It remains to be seen whether the UKIPO and the UK courts will adopt an approach to the examination of patent applications that is consistent with the EPO.
Written by Mark Saunders for ip21
February 2026