UK: Patentability of Computer Programs: EPO Referral Held by Enlarged Board to be Inadmissible

Referral by EPO president held to be inadmissible by Enlarged Board of Appeal - Board of Appeal divergences are legitimate developments of the law

Summary

The eagerly awaited response by the EPO's Enlarged Board of Appeal to the EPO president's referral of questions concerning the patentability of software, case G 3/08, has been issued. The Enlarged Board has ruled on 12th May 2010 that the referral by the president is inadmissible.

While the EPO says on its website that the ruling by the Enlarged Board confirms the current practice by the EPO in this area, observers seeking ground-breaking law on the boundaries of patentability of computer-implemented inventions (CII) may be disappointed. Some may feel that, by finding the referral by EPO president, Alison Brimelow, to be inadmissible, the Enlarged Board has thrown away a valuable opportunity to bring pan-European harmonisation to a difficult legal area. The Enlarged Board has instead used the opportunity to confirm various aspects of specified CII decisions.

As a reason for inadmissibility of the referral, the Enlarged Board of Appeal indicated that the entitlement of the EPO's president to make referrals to it under Article 112(1)(b), EPC, must relate to "different decisions" of the Board of Appeal. The Enlarged Board found that the decisions cited by president, with one exception, did not constitute "different decisions" by the Boards of Appeal. Decision G3/08 indicates that the Enlarged Board understood "different decisions" to mean conflicting decisions and the only conflicting decision acknowledged by the Enlarged Board represented a "legitimate development of the case law". The president was therefore not entitled to make the referral.

While the decision by the Enlarged Board does not provide a new over-arching approach in the determination of patentability for computer-implemented inventions, it does contain guidance on a number of specific issues raised by the referral.

Disparities in approaches to computer-implemented inventions exist across Europe, but the absence of a clear line by the Enlarged Board may be felt more acutely in the UK than in other EPC contracting states, due to the Court of Appeal's criticism of the Board of Appeal and the Court's attempts under both the current and previous EPO presidents to obtain a consistent approach from the EPO. The Enlarged Board, in putting forward the "legitimate development" argument, has clearly rejected this criticism, underlining the difference between the legal tradition in continental jurisdictions and that of precedent-driven English courts.

Where next for computer-implemented inventions?

To view the article in full, please see below:

Full Article

Referral by EPO president held to be inadmissible by Enlarged Board of Appeal - Board of Appeal divergences are legitimate developments of the law

Background

Significant differences have arisen between the approaches by various European jurisdictions to computer-implemented inventions. In short, there is no single approach across Europe to determining if a software-related invention is patentable or not. Since the first such case in Europe in 1984, the Vicom decision, a host of tests has been put forward in various fora, including the contribution approach, the technical effect approach, the whole claim approach, the any hardware approach and others. At various times over the last 30 years each of these tests have been proposed, but, in the opinion of some critics, no definitive test has been identified and the approaches are considered by some to be inconsistent, leading not only to legal uncertainty, but also to considerable economic uncertainty as the asset value of patents moves in accordance with the various rulings.

Article 52 of the European Patent Convention (EPC) provides that:

The controversy hangs on the prohibition in Articles 52(2)(c) and 52(3) EPC which state that programs for computers as such are excluded from patentability. The "as such" is a short, but extremely powerful, qualifier, and has been interpreted by the EPO Boards of Appeal and various jurisdictions to mean that while an invention consisting of pure software on its own is excluded from patentability, an invention which implements a software-based concept may be patentable, depending on how the implementation occurs. The tests for exclusion generally include some test for "technicality", whereby subject-matter which is insufficiently technical in nature is considered non-patentable. With some exceptions (notably by the open source movement), most commentators appear to agree on the approach to this extent, but the current problem hinges on which test to apply to determine if the computer-implementation patentable is sufficiently technical.

Decision G3/08 acknowledges the scale of the controversy and refers to unsuccessful attempts by legislators to clarify the situation with respect to patentability of computer-implemented inventions. The discussions leading to the EPC 2000 included proposals to modify or remove the exclusion, but these proposals were not taken up in order to give the Commission's proposed Directive the opportunity to determine a suitable provision. After modification by the Council, the proposed Directive became the focus of the open source movement and anti-patent groups. Numerous amendments were put forward and the proposal was eventually rejected by the European Parliament in July 2005.

The controversy has continued since. The test applied by the Court of Appeal in the UK, as set out in Aerotel Ltd v Telco, Re Macrossan, [2006] EWCA Civ 1371 and confirmed in Symbian Ltd v Comptroller of Patents [2008] EWCA Civ 1066, is based on the contribution approach originally proposed by the Board of Appeal in the Vicom decision (T208/84). Vicom required identification of a technical contribution to the known art and considered the subject-matter to be patentable if such a technical contribution was present. Similarly, the Aerotel/Symbian test requires the identification of the contribution to the art of the alleged invention, then determination whether the contribution is within the exclusion (eg not pure software per se) and finally checks if the contribution is technical in nature – contributions which are not excluded and are indeed technical may be patentable.

The EPO Board of Appeal stated that it has moved away from the contribution approach and favoured a "further technical effect" approach, first put forward in IBM decision T1173/97 and recently confirmed in Dun Licensing T0154/04 and Gameaccount T1543/06. The Board of Appeal has stressed the distinction between the two separate EPC provisions in A52(1) and A52(2)(3), the former relating to the requirement for inventive step and the latter setting out exclusions to patentability. The motivation behind the Board of Appeal's shift away from the contribution approach and in favour of the technical effect approach aims to keep the test for novelty and inventive step separate and distinct from the exclusion test. The EPO believes that identifying a contribution to the art is part of the test for novelty/inventive step ie an A52(1) test, rather than part of an exclusion test and that the contribution approach, at least in the context of A52(2), is wrong for this reason and wants the test for "technicality", eg the technical effect test, to be independent of any consideration for inventive step.

The main criticism of the Board of Appeal by the Court of Appeal is that the Board has no settled approach to the issue, having over time picked up and abandoned several different tests for exclusion. The Court of Appeal argued that it is therefore unable to follow the Board's decisions and it was in this context that Jacob LJ instigated the referral by the EPO president to the Enlarged Board of Appeal. The first attempt to do so was declined by the then EPO president, Alain Pompidou, on the grounds such a referral was not necessary.

However, his successor as EPO president, Alison Brimelow, decided to make the referral to the Enlarged Board of Appeal. Under Article 112(1)(b) she referred a set of complex questions concerning the patentability of software and computer-implemented inventions on 22nd October 2008, the questions being designed to provide a definitive answer to the seemingly insoluble controversy surrounding the exclusion to patentability in Article 52(2)(c)(3), EPC.

The questions raised in the referral were the following:

For each of the four questions of the referral at least two Board decisions were cited as diverging, along with an analysis of the divergence.

Held

General Admissibility

The Enlarged Board of Appeal considered at length the basis for referrals to it by the EPO president and, in particular, the requirement under Article 112(1)(b) that, in order to ensure uniform application of the law, such referrals may be made in the case of "different decisions" by two different Boards of Appeal.

The Enlarged Board noted, at 7.2.4, that it did not rule on abstract points of law, but only on real issues arising from the cited differing decisions and, at 7.2.5, that "neither the Boards of Appeal nor the President are authorised to consult [the Enlarged Board of Appeal] whenever they so wish in order to clarify abstract points of law." More precisely, the president's right of referral did not extend "to use the Enlarged Board referral as a means of replacing Board of Appeal rulings on CII patentability with the decision of a putatively higher instance", paragraph 7.2.7.
The Enlarged Board referred to the concept of legal development and contrasted judgements in continental civil law with "Anglo-Saxon law, where a decision on an individual case has far greater implications as a precedent".

It added at 7.3.1 that "legal development could not on its own form the basis for a referral" and indicated that "law may not always develop in a linear fashion, and that earlier approaches may be abandoned or modified". "Repeated amendments [of jurisprudence] will be necessary in the course of time if legal solutions that the Boards initially deem correct prove to be mistaken ... and cease to be convincing jurisprudence. Such changes are a normal part of judicial activity and there is no need to speak of different decisions within the meaning of Article 112(1)(b) EPC simply because departures from earlier practice are deemed necessary when homing in on the right solution to a specific case".

The Enlarged Board continued at 7.3.5 "even a radical shift in jurisprudence need not necessarily be construed as a different decision within the meaning of Article 112(1)(b) vis-à-vis earlier case law, provided that the Board corrects itself and – mostly in explicit fashion – declares its earlier practice to be no longer relevant."

In another passage [7.3.6] the Enlarged Board clarified how non-linear development of the law may be legitimate: "Legal development has followed a course which, while not entirely linear, as long as it is transparent, does not justify speaking of different decisions". In making its decision the Enlarged Board said at 7.3.8 that it "must also consider whether the divergent decisions might not be part of a constant development, possibly still ongoing, in jurisprudence on recent patent law issues, in the course of which older decisions have lost their significance and so can no longer be considered in connection with newer decisions. Such putative decisions do not justify presidential referrals..."

The Enlarged Board concluded, at 7.3.8, that the "president has no right of referral simply to intervene .. in mere legal development, if, on an interpretation of the notion of different decisions in the sense of conflicting decisions, there is no need for correction to establish legal certainty."

The "legitimate development of the law" argument was applied only in the case of Question 1, where the Enlarged Board acknowledged there had been an inconsistency between two decisions of the Boards – in respect of all other questions in the referral, the Enlarged Board that no divergence had occurred. In either case the question concerned was not admissible.

Despite judging each question to be inadmissible, the individual inadmissibility analyses by the Enlarged Board do provide some guidance on the issues raised.

Question 1 of the Referral

Question 1 of the president's referral cited decision IBM T1173/97 in which the manner in which a computer program was claimed (eg as a computer program, computer program product or computer-implemented method) was considered irrelevant, the important feature for patentability being whether the claimed matter had technical character. The president contrasted this with Microsoft T424/03, which emphasised the claim form, allowing "a computer-implemented method x" and "a computer program product storing executable code for method x", but not allowing "a computer program for method x". Avoidance of the exclusion, according to Question 1, would become a mere formality, achieved by formulating claims as a computer-implemented method or a computer program product.

On this question, and only this question, the Enlarged Board of Appeal acknowledged, at 10.8 and 10.12, that there was a difference between the decisions cited. The Enlarged Board nevertheless found that there the alleged divergence was a legitimate development of the law (10.12) and secondly, that, no subsequent Enlarged Board decision had, in this aspect, followed T1173/97,(see 10.8.1 and 10.12). T1173/97 remains established case law in other senses, 10.4 and 10.8.1. In the light of the Enlarged Board's general view on divergence or legitimate legal development (see above), Question 1 was therefore deemed to be inadmissible.

The Enlarged Board noted, at 10.13, that explicit mention of the use of a computer or a computer-readable storage medium, will be sufficient to avoid the exclusion, although not necessarily sufficient to meet the inventive step requirement, as set out in Duns T154/04, which represents a "practicable system for delimiting innovations for which a patent may be granted".

Question 2 of the Referral

In her referral question 2, the EPO president cited a divergence between IBM T1173/97 and Hitachi T258/03 (confirmed in T424/03). She indicated that, while a computer program and a computer-implemented method are identical in scope (T38/86), the two decisions T1173/97 and T258/03 treated them differently: the T1173/97 decision treated this as patentable if a further technical effect was demonstrated, while T258/03 treated this as patentable if a technical means was involved.

The Enlarged Board disagreed with the president's basic premise that a computer program and a computer-implemented method are the same thing, arguing at 11.2.4 and 11.2.5 that a computer program refers to a sequence of instructions specifying a method, rather than the method itself. Moreover, expressions such as "a computer loaded with a method" or "a medium storing a method" were meaningless (see 11.2.4), further indicating the non-identity of the terms "program" and "method".

With respect to the two citations, the Enlarged Board ruled that T1173/97 did not imply that features of a computer program product could be treated as method steps. While the Claim 1 of T38/86 was directed toward a method, it did not imply that a computer program could be equated with a method, merely that the claim could be satisfied by the use of a computer program, 11.2.9.

The Enlarged Board concluded that a computer program and a computer-implemented method are not identical in scope and therefore that the divergence cited by the president's referral had not occurred.

Question 3 of the Referral

Question 3 cited Board decisions BBC T163/85 and Mitsubishi T190/94 which required a technical effect to be brought about on a physical entity in the real world, and decisions Henze T125/01 and T424/03 in which technical effects were limited to those occurring in the respective computer programs and independent of any hardware. In her referral the president considered these two positions to be divergent, as the latter decisions would consider effects related solely to the computer programs, eg elementary programming features improving the efficiency of the program, to be patentable, whereas the same subject-matter would fall foul of the requirements set out in former decisions.

The Enlarged Board disagreed with the premises of Question 3, stating (i) that the referral question itself related to individual features (contrary to the Vicom principle that non-technical features could be considered for inventive step analysis), whereas the president's analysis of the divergence referred to claimed subject-matter as a whole, and (ii) that BBC T163/85 and Mitsubishi T190/94 did not in fact state that a technical effect would be required on a physical entity in the real world, only that if this would be sufficient to avoid exclusion.

On the basis that the decisions cited did not relate to the question posed and there was never a requirement to demonstrate a technical effect on a physical entity in the real world, the Enlarged Board concluded that there was no divergence and the question was inadmissible.
Regrettably the supplementary questions 3(b)(c), relating to a necessity to relate an invention to a computer (the "physical entity"), were not addressed by the Enlarged Board.

Significantly, the Enlarged Board appears to confirm a two-part approach: an exclusion test followed by an inventive step test, 12.2.2 . This was the approach taken in the Comvik T641/00 and Hitachi decisions, in which all features of the claim were considered for technical character, followed by a test for inventive step. Although the inventive step test in Comvik/Hitachi involved the ignoring of some (non-technical) features, the Enlarged Board said it was well established that non-technical features cannot be discarded at this stage, 12.2.3.

Question 4 of the Referral

In support of Question 4 the EPO president cited the statement in T1177/97 that "implementing a function on a computer system always involves, at least implicitly, technical considerations" (this being supported by T172/03), and decisions T833/91, T204/93 and Sohei T769/92 in which a programmer's activity, ie writing computer programs, is deemed to be a mental activity and falls within the exclusion to patentability.

The Enlarged Board did not recognize any inconsistency between these positions, holding that a programmer's activity may well be just a mental activity and therefore, to this extent, excluded from patentability. It held that technical considerations were in themselves not sufficient to escape the exclusion - if the technical considerations did rise beyond the level of mere mental activities they would fall within the scope of the exclusion. The Enlarged Board did acknowledge that there is some potential for confusion, arising from Sohei T769/92, that any technical consideration may confer technical character on a claim – this is not the case, as set out in T1173/97, which states that for computer programs technical effects are always present and a further technical effect is required to confer technical character. Thus programming per se may be just a mental activity, but if the programmer has considerations beyond "merely" finding a computer algorithm to carry out the procedure, the corresponding claim may fall outside the exclusion of Article 52(2), EPC.

Once again, the Enlarged Board held that no divergence was identifiable in the cases cited and Question 4 was therefore inadmissible.

Comments

A considerable part of this decision relates to the exclusive right of interpretation by the Boards of Appeal and the Enlarged Board, the Enlarged Board clearly demonstrating its independence. Despite considerable reference in the decision to the ongoing controversy surrounding computer-implemented inventions, the Enlarged Board's ruling may do little to resolve the controversy.

For each of the allegedly differing Board of Appeal decisions cited by the president as giving rise to a conflict in the law, the Enlarged Board of Appeal has stated that such decisions are either not in conflict or, if in conflict, then merely represent a legitimate development of the law. The Enlarged Board indicated also that such development of the law may be non-linear, and it may be appropriate to abandon outdated approaches.

In view of this, observers may question the value of the EPO president's entitlement under Article 112(1)(b) EPC to refer a matter to the Enlarged Board of Appeal. Even in the case of the single referral question for which the Enlarged Board acknowledged a divergence – Question 1 – the Enlarged Board stated that this case was a legitimate development of the law. The Enlarged Board made it clear that "even a radical shift in jurisprudence" may be insufficient for the change in the law to be considered a "different decision" for the purpose of Article 112(1)(b), and the circumstances in which the EPO president's referral may be acknowledged appear now to be narrow.

In view of the inadmissibility finding, the Enlarged Board did not provide an over-arching guidance on patentability of computer-implemented inventions. Nevertheless, the inadmissibility arguments put forward by the Enlarged Board do provide some guidance on a number of discrete issues:

1. The Enlarged Board appears to consider that the form of the claim is indeed relevant, formulations such as "a computer-implemented method x" and "a computer program product storing executable code for method x", being potentially patentable, but the form "a computer program for method x" falling foul of the exclusion. In this sense T1173/97 appears to have been replaced by T424/03, as a legitimate legal development.

Explicit mention of the use of a computer or a computer-readable storage medium, will be sufficient to avoid the exclusion, although a claim will have to satisfy the inventive step requirements set out in the decision Duns T154/04.
The Enlarged Board described Duns T154/04 as "a practicable system for delimiting innovations for a which a patent may be granted."

2. The Enlarged Board held that "a computer program" is not the same thing as "a computer-implemented method" and that the two decisions, T1173/97 and T258/03, which related respectively to these concepts, were justified in treating them differently with respect to the exclusion.

3. Demonstrating a technical effect on a physical entity in the real world would be sufficient to avoid the exclusion, but there is not per se a requirement to do so.

4. The Enlarged Board confirms a two-part approach when determining the patentability of a computer-implemented invention, consisting of an exclusion test followed by an inventive step test (see reference to Duns 154/04 above).
The Enlarged Board confirmed that the implementation of a function on a computer system always involves, at least implicitly, technical considerations, but such technical considerations may still represent only a mental activity and still require a further technical effect, as set out in T1173/97, to confer technical character.

For observers in Anglo-Saxon jurisdictions, G03/08 is a clear demonstration that the EPC's ultimate interpretation arbiter sees no place for the common law doctrine of binding precedent and is unconcerned by significant changes in the law of patentability over time. By contrast, the Court of Appeal in London, while it may appreciate the desirability of pan-European harmonisation, may feel vindicated in its complaint that the decisions of the Board of Appeal are not settled, G03/08 showing that the Enlarged Board of Appeal is manifestly undisturbed by this.

The clash of legal cultures, exposed by G03/08, seems set to continue, with both UK courts and the Boards of Appeal deciding on the patentability of computer-related inventions in the UK. The controversy is set to continue and, in the absence of any renewed legislative initiative, for which there seems little political appetite, it will be interesting to see if this can be resolved by any future European Patent Court.

This article was written for Law-Now, CMS Cameron McKenna's free online information service. To register for Law-Now, please go to www.law-now.com/law-now/mondaq

Law-Now information is for general purposes and guidance only. The information and opinions expressed in all Law-Now articles are not necessarily comprehensive and do not purport to give professional or legal advice. All Law-Now information relates to circumstances prevailing at the date of its original publication and may not have been updated to reflect subsequent developments.

The original publication date for this article was 24/05/2010.