Comparison of co-pending UK and European patent applications for computer-implemented inventions

Do you get the same result when prosecuting a UK patent application and a European patent application for the same computer-implemented invention (all other things being equal)?

This goes to the heart of the issues currently being considered by the UK Supreme Court in Emotional Perception AI Limited v Comptroller General of Patents and is of the upmost importance to those looking to obtain and enforce patents in the UK for inventions relating to AI, quantum computing and other areas of computing. One of the questions being posed is whether there is a material divergence in practice between the UK and EPO approaches to the assessment of computer-implemented inventions (CIIs).

The UK IPO thinks not, but many patent practitioners in the UK beg to differ. This article takes a closer look at this issue.

The IPO view

In the written case made on behalf of the Comptroller General of the UK Intellectual Property Office (IPO), the IPO remarked that no cases had been given of co-pending patent applications which demonstrate a material difference between the UK approach – the Aerotel/Macrossan test – and the EPO approach set out by the Enlarged Board of Appeal in G1/19 Pedestrian Simulation. In particular, the IPO pointed to an absence of examples in which the UK application had been refused for excluded subject matter, but the European counterpart for the same invention had been allowed by the EPO.

Searching for evidence

The task of finding co-pending applications, however, is not straightforward. It is rare for an applicant to prosecute both UK and European patent applications. There are several reasons for this. One reason is that there is seldom a need to obtain both a UK patent and a European patent (UK) for the same invention and, indeed, Section 73(2) of the Patents Act prohibits such double patenting. Another reason is that if there are two applications and prosecution of one case is proving easier, then the applicant is likely to drop the other, more arduous case.

This assumes that an applicant is actively seeking to obtain protection at both the IPO and EPO. This is not normally the case. Most UK applicants abandon their UK priority application if they then file a direct European patent application or an international application (with a view to pursuing a Euro-PCT application later). There is simply no reason to continue with the UK application.

Moreover, in many cases the applicant withdraws their UK patent application or abandons it without the application being published, which crucially means that the file is not laid open to public inspection. This last point is important since it means that it is impossible to investigate those cases without the applicant voluntarily publishing the file. Lastly, some applicants and UK patent attorneys do not file UK applications except for the strongest, most “technical” CIIs due to the perceived difficulty of obtaining protection for CIIs at the IPO.

This means that the pool of co-pending applications which can be meaningfully compared is reduced, making the task of identifying and analysing co-pending applications even harder.

Abandoned, withdrawn or refused: does it matter?

It should be noted that the Comptroller referred to finding an example of a UK application that had been refused, as opposed to being abandoned or withdrawn.

The pool of UK applications which are actively refused is small. The number of cases which are abandoned, on the other hand, is much higher. As explained earlier, if an applicant is having a hard time at the IPO and it has a co-pending EP application which is progressing more smoothly, then it is likely to abandon the UK case. This can be done, for example, by not filing a response to an examination report and letting the compliance period expire (usually at four-and-a-half years from the filing date for an initial UK application). Thus, the application is abandoned but not refused.

Focussing on such cases, it is possible to perform searches using publicly available patent databases which reveal several UK patent applications that the IPO has refused to search (because the invention is excluded from patentability) or raised an exclusion objection and yet resulted in a granted European patent. Six examples are listed below.

These examples illustrate that the differences in practice between the UK and EPO approaches to patentability do lead to different results. This disharmony is harmful to applicants. For example, if an exclusion objection is raised early in the application process, it can put off an applicant from pursuing the UK application and, further still, dissuade them from filing applications outside the UK. For example, the IPO can issue an abbreviated search report stating that no meaningful search can be performed. Thus, no prior art is cited. If the search report is accompanied by an examination report which raises an exclusion, it will usually contain only a brief analysis of the invention and a conclusion which pigeonholes the invention (most often) as a “computer program as such”.

Faced with no prior art and a stark “this-is-fundamentally-unpatentable” rejection, is it any surprise that an applicant – particularly if inexperienced or on a tight budget – chooses to give up on patent protection?

Cui malo

So, who suffers when there is a divergence in approach?

The greatest impact is likely to be felt by UK-based research institutions, start-ups, scale-ups and high-tech SMEs who are filing patent applications for the first time. These entities tend to file first in the UK and decide what to do next based on the feedback from the IPO. It is precisely those entities that would benefit hugely from a proper assessment of their priority applications at the IPO.

Examples

The following are just a few examples of co-pending applications for which exclusion from patentability objections were raised on the UK application and yet resulted in a granted European patent.

GB 2012964.9 and EP 21188902.7

This case relates to a searchable encrypted database and to a method of searching the database.

The IPO examiner did not perform a search and issued an abbreviated search report which stated that the claimed invention was excluded from patentability. Following the Aerotel/Macrossan test, using the independent claims as a starting point to identify the alleged contribution and considering the guidance in the form of signposts provided in AT&T Knowledge Ventures LP and CVON Innovations Ltd v Comptroller General of Patents, the examiner took the view that the method involved simply returning search results to a user and, therefore, was simply the presentation of information as such. The examiner also concluded that it was a computer program as such since the entire method was carried out on a computer.

The applicant amended the claims, and an examination report was issued maintaining that the invention was excluded from patentability. The application process ran its course and was terminated at the end of the compliance period.

A direct European patent application was filed claiming priority from the UK application with the same claim 1. The EPO carried out a search and in the accompanying search opinion, the examiner considered that claim 1 lacked an inventive step in view of prior art cited in the search report. The EPO examiner did not flag any features as being non-technical and did not think that the claimed method was being used for a non-technical purpose. There was no mention of COMVIK (or G1/19). In short, the claims were examined in the same way as any claim containing technical features would be examined at the EPO. The claims went through three rounds of amendments, and a European patent was ultimately granted.

Granted claim 1 was narrower than the claim as filed. Nevertheless, the claim was still directed to a method of providing a searchable encrypted database and for searching that database. The comparison of the two cases illustrates the rigid, harsh approach of the Aerotel/Macrossan test.

GB 1901369.7 and EP 20155002.7

This case relates to data encoding of a time series.

The IPO examiner did not perform a search and gave his reasons in the accompanying examination opinion. Again, the examiner followed the Aerotel/Macrossan test: he identified the alleged contribution to be a particular processing and encoding of particular data and asked whether the contribution fell solely with excluded subject matter using the AT&T signposts. The examiner concluded that the contribution fell within the computer program exclusion. An abbreviated examination report was issued repeating the objections raised in the earlier examination opinion. The claims were not amended and the compliance period expired.

A direct European patent application was filed with the same claim 1 as the UK priority application. The EPO carried out a search and, in the accompanying search opinion, the examiner considered that claim 1 lacked novelty in view of the cited prior art. The examiner, however, also considered that the claim lacked an inventive step over a notoriously known general purpose computer, following the COMVIK approach (there is no mention of G1/19 since that decision was handed down 6 months later).

The claims were amended twice, and a patent was granted with some additional features, namely specifying a request including a threshold criterion and returning a dataset meeting the requested threshold criterion.

The applicant argued that the features of claim 1 produce the effect of compressing the input time-series data and then returning a subset of this original data that meets a desired level of similarity, fidelity or accuracy, in response to a corresponding request.

GB 1912482.5 and EP 20193457.7

This case relates to a software license distribution system.

For the GB application, the search was contracted out by the IPO to the EPO which carried out a search citing (among others) two category X documents. The IPO then issued an examination report, applying the Aerotel/Macrossan test and using the AT&T signposts. The examiner considered that the invention related to a computer program as such and method of doing business as such and, thus, was excluded from patentability. The applicant argued against the objection and another examination report was issued maintaining the objection. The applicant did not reply and so the application process terminated at the end of the compliance period.

A direct European patent application was filed with an identical claim 1. Unsurprisingly, the same prior art was cited (since the EPO had carried out the search on the UK case). The EPO examiner, however, did not have any issue with features being non-technical or the claim being directed to a non-technical purpose. After two rounds of amendments, the European patent was granted.

GB 1810910.8 and EP 19736459.9

This case relates to providing access to at least one password protected device via a password management system.

The UK application followed a familiar path: the IPO examiner did not perform a search and issued an abbreviated search report in which it stated that the claimed invention was excluded from patentability.

An international application was filed with the same claim 1. The EPO (as the International Searching Authority) carried out a search and examined claim 1 without needing to apply the COMVIK approach.

The application entered into the European phase as a Euro-PCT application. Although the claims were amended to address lack of unity and lack of clarity objections, and to introduce reference numerals, the EPO granted a patent based on claim 1 with minor changes.

Thus, the EPO granted a patent with a claim that the IPO examiner considered to be excluded from patentability.

GB 1706442.9 and EP 17718727.5

This case relates to data processing.

In this example, a search was performed by the IPO. It was, however, accompanied by an examination report which considered that the contribution related to essentially abstract data handling and fell wholly within the computer program exclusion. The examiner persisted with the objection and the application process terminated at the end of the compliance period.

An international application was filed claiming priority from the UK application. The application entered into the European phase as a Euro-PCT application, and the EPO granted a patent with claim 1 unchanged in substance.

Thus, once again, the EPO granted a patent with a claim that the IPO examiner considered to be excluded from patentability.

GB 1903870.2 and EP 17787563.0

The final example relates to building control systems.

This case is unusual in that the UK application procedure ended with a decision by an IPO hearing officer refusing the application. The hearing officer applied the Aerotel/Macrossan test and considered the AT&T signposts and concluded that the claimed invention related to a computer program as such.

The Euro-PCT application was allowed with a narrower claim. In its capacity as the ISA, the EPO initially considered claim 1 to lack an inventive step on the grounds that it did not produce a technical effect. After some amendments, the EPO granted a patent.

Therefore, this was an example of precisely the type of case requested by the Comptroller.

Conclusion

These cases are not the only examples showing different outcomes at the IPO and EPO: the searches revealed several other examples too.

It should be noted that no judgement is being made here as to whether these cases are “good” examples, or in other words, clear-cut cases that should (or should not) have been allowed. And, of course, it may be that the IPO’s assessment was correct, and the EPO should not have granted a patent.

What these cases do show, however, is that the outcomes at the IPO and EPO differed and that the UK approach to the assessment of the computer-implemented invention seemed to make it easier for the IPO to object to the claims.

Interestingly, in some cases, EPO examiners simply went ahead and examined the claims without questioning the technicality of the claim features and without feeling the need to apply COMVIK. This suggests that the EPO might have a different view on what constitutes a technical feature or a technical purpose in some areas.

Having different approaches to assessing computer-implemented inventions at the IPO and EPO only makes it more likely that there will be different outcomes at the two offices. This is clearly undesirable, and measures must be taken – if not by the Supreme Court, then by the IPO – to address this issue.