Software Patents: Another fine mess…

February 23, 2007

In the October/November 2006 issue, Colin Hulme argued that until the IP laws concerning the protection of software have been harmonised, both sides in the software patent debate would lose out. I certainly agree, but wonder whether there could ever be a fix. The recent decision of UK Court of Appeal in Aerotel v Telco Holdings Ltd et al, and Macrossan’s Application [2006] EWCA Civ 1371 (actually two decisions in one) shows that, despite the wording of the UK and European patent law being substantially the same, the UK and European Patent Offices are incapable of interpreting it in the same way.

Aerotel and Macrossan was intended, as such decisions often are, to clarify the position in the UK. But for those with an eye to Europe it has merely added confusion, and made the UK an extremely unfavourable jurisdiction for obtaining patent protection for software related inventions. Although the question of whether software patents are a good thing or not for the UK software industry remains hotly debated, what all agree on is that uncertainty is never beneficial.

The Present European Patent Office Position

Although Hulme has categorised the situation in Europe as ‘in utter disarray’, the European Patent Office (EPO) has taken a fairly consistent view on software and business methods for the past several years. The EPO Board of Appeal (BOA) decisions may appear inconsistent when taken at face value, but the BOA is not bound by its former decisions, only those decisions that have issued from the Enlarged Board. The apparent inconsistency therefore stems from the older decisions of the BOA being superseded by newer decisions that reflect the updated EPO view, and it is entirely removed when one looks at chronology and the direction the decisions point in. Specifically, there has been a refinement of the EPO’s position, leading to the current ‘technical solution to a technical problem’ test that is used to decide whether a patent can be granted for potentially-excluded subject matter.

The Current European Test

In practice, the European test turns out to be less stringent than the current test applied in the UK. As Hulme noted, following the Hitachi decision (T0258/03), the EPO interprets the exclusion to computer software ‘as such’ to render unpatentable only  ‘purely abstract concepts devoid of any technical implications’. This does not however open the way for patents to be granted for any and all computer software. Although the EPO regard a programmed computer as a technical entity not an abstract one (so a possible ‘technical solution’ for the purposes of the test) they nevertheless still reject most software-related patent applications on the grounds that, once the computer is programmed, no technical problem is solved or that, if a technical problem can be identified, the software is merely an obvious implementation. They have therefore adopted a slight-of-hand that allows them to reject software not under the excluded subject matter exemption of Article 52 of the EPC, but under the inventiveness test of Article 56. Though the legal basis of the objection has changed, the focus is still on identifying a technical solution to a technical problem. I believe that this test is clearer, more certain, and more workable than that put forward in Aerotel and Macrossan.

The UK Test set out in CFPH LLC

In contrast with Europe, certainty has been lacking in the decisions of the UK Patent Office for several years, and the interpretation of what is excluded subject matter has grown progressively narrower. Prior to Aerotel and Macrossan, the UK test for potentially-excluded subject matter was set out in CFPH LLC [2005] EWHC 1589 of the UK High Court, and comprised a two-step test proposed by the Deputy Judge. The test involved identifying the ‘advance’ made by the invention, and confirming that the ‘advance’ fell within the definition of ‘an invention’ in the sense of Article 52 of the EPC. Despite specifically referring to the European Patent Convention (EPC) in the test, and despite the Deputy Judge explicitly using the European ‘technical solution to a technical problem’ approach as a way of interpreting both the existing UK test for excluded subject matter (endorsed in Merrill Lynch’s Application [1989] RPC 561 (CA)) and the facts of the case before him (see the CFPH LLC decision at [118]), when the new wording was adopted by the UK Patent Office, it was misinterpreted to mean that consideration of whether the software was doing something ‘technical’ was no longer necessary. This was contrary to the European position, and it is not clear if the UK Patent Office’s error has yet been fully redressed.

The New UK Test

The new UK test, confirmed in the Aerotel and Macrossan decision, is based on the formulation put forward by the UK Patent Office, and focuses on identifying a contribution in a field that is not excluded, rather than on identifying a technical solution to a technical problem. Having identified the contribution made by the invention, the test is to:

(3) ask whether it falls solely within the excluded subject matter;
(4) check whether the actual or alleged contribution is actually technical in nature.

Although, the last step refers to a “technical” contribution, the emphasis of the test is on the previous step (the fourth step is regarded as an unnecessary but useful check).  The test has subsequently been applied in Capellini & Bloomberg v The Comptroller-General of Patents [2007] EWHC 476 (Pat), where the judgment was handed down by the High Court on 13 March 2007.

The test is regarded by the UK Patent Office as giving the same result as the European test in ‘nearly every case’, and yet I believe that it is significantly narrower. Why is the test narrower? A good example is software for compiling programs: the EPO regard a compiler producing programs that utilise system resources on a computer as a possible technical solution to a technical problem (the UK Patent Office similarly had no objection once upon a time) because the performance of the computer running the resulting program is generally improved as a result. Now in the UK, however, a compiler producing a more efficient program has merely made an advance in the field of computer programming and so is excluded.

Another example is all claims that are directed specifically to ‘a computer program’ itself, rather than to computer programs defined as ‘methods’ or programmed ‘apparatus’. In Europe, claims to a computer program are patentable providing that a similar claim drafted as an apparatus or as a method would also be patentable (most European patent applications for software contain both apparatus, method and computer program claims). The UK Patent Office have indicated in their recent practice note (Patents Act 1977: Patentable subject matter Notice:, however, that they no longer believe claims drafted to ‘a computer program’ are viable.

If you were a large computer software manufacturer, would these differences affect whether you would choose to site your office in the UK?

More Confusion?

The new test put forward in Aerotel and Macrossan is not just narrower, it is also more confusing. In the Court of Appeal decision, two inventions were considered. In Macrossan the invention was a computer programmed to obtain information from a client, and prepare legal documents; in Aerotel, it was a method and system for providing a prepaid phone service. Although one invention was found to be patentable and one was not, both inventions could be said to be an advance in excluded subject matter, and therefore fail the third part of the test.  Although it is a European perspective, neither of the inventions address a ‘technical problem’: the first is arguably automating what solicitors do, and the second provides an implementation of a prepaid phone call business. If there is no technical problem, isn’t the claimed solution an obvious one?

Aerotel was eventually allowed because it claimed a telephone system having a ‘special exchange’, while Macrossan was refused because no new hardware was provided above a conventionally programmed computer. The confusing thing is that the Aerotel ‘special exchange’ could be (and probably is) a conventionally programmed computer like that used in Macrossan, thus reducing the difference between the inventions to a semantic one. Given that the contribution in Aerotel is a pre-payment system, one can’t help thinking that Aerotel succeeded where Macrossan did not merely because their invention related to the telecommunications field rather than document preparation.

In practice, the test set out in Aerotel and Macrossan is regarded as preferable by the UK Patent Office because it avoids having to rely on the meaning of ‘technical’. The EPO approach for example is often criticised because of difficulties in identifying what is ‘technical’ and what is not. However, I believe it is no easier to identify the nature of the ‘actual or alleged contribution’ required under the new UK test.


Such difficulties would perhaps be academic if it was not for s 130(7) of the Patents Act 1977. The interpretation of the exclusion of software and business methods was, prior to the introduction of that Act, intended to be the same in the UK as in Europe. Section 130(7) states that ‘as nearly as practicable’ the wording of the UK Patents Act and the European Patent Convention that excludes ‘software as such’ from patent protection should have ‘the same effect in the United Kingdom as the corresponding provisions of the European Patent Convention (EPC)’ (the EPO BOA decisions are based on the wording of the EPC, which as noted above is substantially identical to the wording of the Patents Act).

Despite that, the interpretation is clearly not the same, and the recent approach put forward by the UK Patent Office and confirmed in Aerotel and Macrossan has not improved the situation. This is more than a little unsatisfactory, as patent protection in the UK can be obtained either by having your patent application examined at the UK Patent Office or by having your patent application (a European application designating at least the UK) examined at the European Patent Office. The European route to protection is more expensive, which means that if the results of these two routes is not the same, then UK patent applicants who cannot afford to file a European patent application are at a disadvantage – not just over their UK competitors who can afford to and who do apply for a European patent, but also over similar competitors from around the world.

The request of the Court in Aerotel and Macrossan to clarify the EPO position is welcome, as in responding the EPO will have the opportunity to put forward a definitive statement of their position, in a way that was blocked by the proposed Directive of Computer Implemented Inventions. The UK Patent Office and courts will also be able to apply that definitive statement.

The Aerotel and Macrossan judgment was somewhat unusual as it contained an invitation to the EPO to clarify how the exclusions to patentable subject matter should be interpreted. This invitation was certainly welcomed in the UK as, if the EPO had taken it up, the EPO could have put forward a definitive statement of their position in a way that was blocked by the proposed Directive of Computer Implemented Inventions, and that the UK Patent Office and courts would subsequently have been able to apply. However, the President of the EPO has refused, at least in part because he did not see sufficient discrepancy in the decisions of the EPO to warrant the investigation. Although, I agree with the President’s analysis on this point, it is regrettable that the opportunity was missed to offer further clarification and cement the present position for the benefit of all the European states that contract to the EPC.

Until this happens, UK software developers are faced with a UK Patent Office that is more likely to throw out software related patent applications than its European and American counterparts. While an expert practitioner can often push something over the line, extra preparation and care is required. Given the ease with which software can be distributed, either electronically or by traditional posting, and an increasing focus on adequate protection for IP that is necessary to secure funding and sometimes sales, UK software developers are competing on the world stage at a disadvantage.


The question of whether patents should be granted or not for software related inventions is a valid one, but unless the playing field is levelled in other jurisdictions, the current approach of the UK Patent Office against software patents has merely left us in a fine mess…

Nick Reeve is a Partner at Reddie & Grose, the patent and trade mark attorneys.