G 1/23 – But what if the prior art is (use of) a software product?


What constitutes prior art is a fundamental question in patent law.

In T 438/19 (the case which led to G 1/23) some questions have been referred to the Enlarged Board of Appeal to seek clarification on the enablement test for prior use of a product. The questions relate to the circumstances under which the prior use of a product may be excluded from the prior art for lack of enablement. In short, the referral deals with the question of whether a marketed product is excluded from the prior art if its composition could not be analysed without undue burden. More particularly, the referral asks whether the non-enabling prior use of a product excludes the composition of the product from the prior art or whether it also excludes the product per se.

This is of significance, because if the product per se is excluded it cannot be used as a starting point for an inventive step analysis. The referral also seeks clarification over the definition of enablement with respect to the prior use of a product, and particularly whether “enablement” requires that the skilled person be able to fully analyse and reproduce the exact same product. One of the questions refers to a much earlier but very important decision G 1/92 (of 18 December 1992) which dealt with a similar question.

Once the questions of the referral G1/23 have been published, the Enlarged Board of Appeal have given the public the opportunity to comment on these questions or to suggest answers to these questions in the form of an amicus curiae brief. Forresters (under the lead of Michael Fischer) have seized the opportunity to submit an amicus curiae brief not to suggest answers to the questions but to make everybody aware of the fact how broad the referred questions are and how broad the answers given by the Enlarged Board of Appeal could be.

It must not be forgotten that the present referral and also the decision G 1/92 come from a Board of Appeal dealing with cases from the field of chemistry. However, G 1/92 clearly says that its findings apply “mutatis mutandis to any other product”. Looking back in history, one can see that G 1/92 has indeed been cited in cases which did not deal with chemical products, but for example in cases where the prior art was (use of) a software product (e.g. T2440/12). Depending on how broad the answers in G 1/23 will be formulated, it seems natural to assume that G 1/23 will be cited in cases where the prior art is not a chemical product, but a software product.

This leads to the question whether analysis of a chemical product can be equated with the analysis of a (compiled) software product which is available on the market in order to find out its internal functioning (e.g. underlying algorithm). It should be borne in mind that while analysis of a compiled software product means the de-compilation or reverse-engineering of this product, decompiling a software product is allowed in the EU only under special circumstances and for this reason alone cannot be expected by the person skilled in the art. By contrast, in chemistry, analysis of a chemical product typically belongs to the proficiency of the skilled person in this field and is not restricted by any law.

Therefore, Forresters’ amicus curiae brief hopefully helps to make everybody aware of the breadth of the questions referred and to make the Enlarged Board of Appeal conscious of the breadth of the answers that they will give.

Michael Fischer