EPO Case Law on Patentability of Software Inventions From the 80ies Until Now

Fernando Pessoa contemplating on patentability of software inventions.

Back in September 2010 the EPO-sponsored 15^th European Patent Judge’s Symposium took place in Lisbon, where the judges discussed developments in European patent law. The symposium was attended by 120 patent judges from Europe and guests from the USA and Japan. A big issue of course was the EU patent and the related litigation system, but the related discussion is now meaningless in view of the recent developments in this regard. But there were other interesting topics discussed as well.

For instance, in a working session on “Patentability of Computer implemented inventions”, Mr. Dai Rees – Chairman of an EPO appeal board and member of the Enlarged Board of Appeals that issued the G3/08 opinion – outlined the developments of the respective EPO case law. His contribution is documented in an article in the conference proceedings: Special edition 1/2011 of EPO Office Journal (PDF, 5 MB), pp. 93 to 102. The below summary reflects my personal observations and understanding of Mr. Rees presentation and also contains some individual flavours and thoughts.

Relevant Provisions: As a reminder, the provisions of the EPC to be considered for computer programs or technicality issues not only involve Art 52 EPC (“programs for computers as such“), but also Articles 18, 19, 21, 22 and Rules 42(1)(a), (c), 43(1) and 44 EPC, which use the term “technical” in different contexts. Further, in the EPC 2000 revision, Art 52(1) EPC was clarified in that patents shall be granted for inventions “in all fields of technology”, which, however, was regarded as a clarification rather than a substantive amendment of the Convention.

Early Case Law: One of the early patentability tests applied was the so called “contribution approach”, as adopted e.g. in T 208/84 (VICOM), T 26/86 (Koch & Sterzel), or T 6/83 (Data processor network/IBM). According to this approach, a claim was refused as non-patentable subject-matter, if this claims contribution to prior art – by equally considering all of its features – exclusively lies in matters of Art 52 (2) EPC, such that the claim was considered having no “technical effect”.

According to the contribution approach, patentability was measured by the technical effect of the invention as compared to that of the prior art. For example, in T 208/84 and T 26/86 the claims differed from prior art by an implemented program that improved image quality and enabled longer maintenance intervals, respectively, which provided the required technical effect. If a technical effect different from that of prior art was not achieved, there was no “technical contribution” and the claim was considered non-patentable, like in T 52/85 (Listing of semantically related linguistic expressions/IBM), T 158/88 (Character form/Siemens) or T 603/89 (Marker/Beattie).

However, in a number of early cases (e.g. T 38/86 and T 65/86) it was already concluded the inventive contribution must lie in a “field of “technology” to prevent lack of an inventive step, thus paving the ground for today’s approach to patentability of software inventions.

Modern Case Law: T 1173/97 (Computer program product/IBM) clarified that program claims causing a technical effect upon execution are patentable without the necessity of claiming a carrier medium. This case also marked the end of the “contribution approach”, since it was concluded that the “technical effect” should not be decided in comparison with prior art but in an a priori assessment and that the “technical effect” requires something beyond the normal physical and electrical effects common to all executed programs, thus postulating a “further technical effect” requirement.  The essential finding that programs are potentially patentable was then extended to functional data in general in T 163/85 (Colour television signal/BBC).

Now that the “technical contribution” has been moved to the assessment of inventive step by T 1173/97, a series of three decisions developed the “any technical means” approach for assessing the patentability requirements of Art 52 (2), (3) EPC. In T 931/95 (Pension benefit system) it was decided that a computer set up to carry out a program is potentially patentable regardless of whether the program was patentable itself. T 258/03 (Hitachi) extended this conclusion to methods of operating a computer, provided the method involved the use of technical means (the computer). Finally, T 424/03 (Microsoft) concluded that a claim directed to a program on a data carrier is potentially patentable as well, regardless of the fact that it also claimed the program. Consequently, according to the “any technical means” approach a claim can avoid exclusion on the basis of Art 52 (2) EPC by including a reference to an apparatus, e.g. a computer.

The problem-solution approach: The above-sketched development of modern case law, especially the suggestion in T 1173/97 that the “technical contribution” is an inventive step consideration and the observation in some early cases (e.g. T 38/86 and T 65/86) that the “inventive contribution” must lie in a “field of “technology”, almost naturally lead to the problem-solution approach as developed in T 641/00 (COMVIK) and T 258/03 (Hitachi) and theoretically justified in T 154/04 (Duns).

This approach nowadays is the crucial test to differentiate between a technical contribution implementing a non-technical concept (e.g. a business method) and an inventive contribution in a technical field (e.g. an embedded control software). Its general idea is that only the technical features of a claim may be taken into account for assessing inventive step, while the non-technical features form a basis for formulating the underlying problem, with the effect that the non-technical features may render the technical solution obvious.

The Enlarged Board’s G 3/08 opinion: In October 2008 the following four questions were referred to the Enlarged Board of Appeal (EBA) under Art 112 (1) (b) EPC by former EPO President Alison Brimelow (see FAQ).

1. Can a computer program only be excluded as a computer program as such if it is explicitly claimed as a computer program?
2. Can the exclusion be bypassed by tying the claimed invention to a piece of hardware?
3. Must a claimed feature cause a technical effect in the real world in order to contribute to the technical character of the claim?
4. Does the programming of a computer necessarily involve technical considerations?

In May 2010 the Enlarged Board issued its opinion. Even though the referral was considered inadmissible, the EBA commented on the four questions as obiter dictum and came to a number of substantive conclusions regarding patentability of software inventions.

Questions 1 and 2 asked for the role of the claim category and the claims as a whole, respectively, and related to the “any technical means” criterion (saying that it is sufficient to escape the Art. 52 EPC exclusion if any technical means is claimed) as specified in T 424/03. Even though there was found a slight inconsistency between the “any technical means” approach and T 1173/97, which rather stated that it is irrelevant for an exclusion from patentability whether the computer program is claimed directly or a data carrier having the program stored thereon, the EBA did not qualify this as a “divergence” within the meaning of Art. 112 (1) (b) EPC. The “any technical means” approach thus was confirmed.

Question 3 asked whether an individual feature of a claim must have “a technical effect on a physical entity [...] in order to contribute to the technical character of the claim” to be considered for inventive step. The EBA dismissed this question, since relevant case was seen fully consistent in the sense that all features of a claim are equally considered without any weighting or essentially assessment. As to the “technical effect on a physical entity”, the EBA pointed out that individual features relating to excluded subject-matters may nonetheless contribute to the technical character of a claim and thus cannot be discarded for inventive step considerations, as already stated, although within another context, in T 208/84.

Question 4 made the point that certain decisions regarded programming as necessarily involving technical considerations while others found that programming was a mental activity and thus excluded under Art. 52 (2) EPC. The EBA argued that a claim to a mental act could be rejected even if that mental act involved technical considerations – and dismissed this question as well, while simultaneously admitting that there at least has been some potential for confusion, since some decisions, e.g. T 769/92 (Sohei), might have given the impression that technical considerations by the inventor were sufficient to guarantee the technical character of a claim. In order to establish consistency with T 1173/97, the EBA took the position that “further technical considerations”, i.e. considerations beyond that of formulating an algorithm executable on an unspecified computer, are required to lift a claim over the hurdle of Art. 52 EPC.

Conclusion: As can be concluded from the above, the legal development in this crucial question did not always go linearly and some approaches were abandoned or modified throughout the years. However, since the G 3/08 opinion basically affirmed the status quo, i.e. the pragmatic problem-solution approach as sketched in T 641/00 (COMVIK) and T 258/03 (Hitachi), case law has reached a rather stable situation, providing legal certainty for applicants of software inventions. Thus, in this particular question, Europe appears to be one step ahead of the US and other important patent jurisdictions in the world.

[Photo (C) 2006 by "Shadowgate" via Flickr under the terms of a CC license]