New Amendments: Essential Items

Justification

The concept of technical contribution has pervaded the discussion about the directive and generated great confusion and therefore to some extent deserves to be clarified. While intuitively and in the subjective belief of most commentators the technical contribution appears to be related to the question of patentable subject matter (Article 52 EPC), the EPO used the term as a means of abolishing the subject-matter test by mixing it with the non-obviousness test (Article 56 EPC) in obscure ways, which national courts and ministerial patent officials have found difficult to follow. A similar amendment was adopted in first reading by the EP. This amendment adds some ideas of the Council such as that of subtracting the prior art from the claimed object.

This amendment is very similar to what was approved in JURI. It corrects one error in the second sentence however: the JURI version states that the technical contribution is the set of features which is claimed to differ from the state of the art. This implies that all features not part of the state of the art are by definition technical, which is of course not necessarily the case.

Justification

The Council's proposal is inconsistent here. In its article 2b, the Council says that a technical contribution must be new and non-obvious (= involve an inventive step). In this article the Council says that for an invention to be non-obvious, there must first be a technical contribution. This amendment resolves the contradiction by bringing this article in line with Article 2b, which represents the common sense of patent law in Europe as used by most national courts today. It was also used by the EPO before 2000, when the "Controlling Pension Benefits System" decision brought the confusion, apparently in a hectic attempt to create new rules for the planned directive.

This amendment is a simplified version of the amendment adopted in JURI. It also fixes that amendment's second sentence which indicates that the contribution can consist entirely out of non-technical features.

Justification

Article 52(2) EPC states that programs for computers, along with aesthetic creations, mathematical methods, business methods et al, are not inventions in the sense of patent law. Art 52(3) limits the exclusions to subject matter and activities as such. There has been much dispute about how article 52(3) should be applied to 52(2). While it is a good idea to transfer Art 52 EPC into EU law, care should be taken to transfer not only the words, but also to resolve the ambiguities and thereby achieve harmonisation and clarification.

Justification

The Council's version is tautological and implies that business methods are patentable inventions when they "produce a further technical effect", i.e. when they fulfill a condition which the European Patent Office, which invented this rhetoric in 1998, has admitted to be meaningless.

Since computers are well known, the presence of a computer can of course not by itself constitute a technical contribution. The question is whether the presence of a computer in combination with an improved algorithm can constitute a technical contribution. By failing to pose this question, the Council seems to imply a positive answer.

The distinction between "business method" and "invention which implements a business method" is a common technique for circumventing Art 52 EPC.

The question of how the "invention" is expressed has never been relevant, nor has the distinction between more or less human-readable descriptions of programs. This subsentence serves no regulatory purpose, apart from insinuating that Art 52(2)c EPC should be interpreted in a way that makes it meaningless.

The sentence

"inventions involving ... business methods ..., which implement ..., shall not be patentable."

is syntactically ambiguous but probably means that "business method inventions" are patentable, if they "produce a further technical effect".

The term "normal physical interactions between a program and a computer" means about as much as "normal physical interactions between a recipe and a cook".

In 2000, EPO itself has criticised this wording and explained that it was merely a wordplay temporarily used in the IBM decision of 1998 in order to circumvent the European Patent Convention, in anticipation of a change of law that would render it unnecessary:

http://www.european-patent-office.org/tws/appendix6.pdf:

This amendment fixes the errors while trying to stay as close to the original wording as possible.

Justification

It is contradictory to say that computer programs can not be inventions and yet can be objects of patent claims. This is why the Commission also did not allow program claims in its original proposal.

The condition after "unless" in the Council version is always true, provided that the patent application was properly drafted. The Council amendment appears to pretend that, while it is allowing program claims, it really only means to allow process claims, and that the program claims are really included -- in defiance of the logic of the patent system -- as a kind of additional enforcement tool in cases where the inventor did not invent software as such but some kind of technical process beyond software.

This however would mean that the inventor could obtain a monopoly on something which he did not invent and which, in most cases, will be neither new nor non-obvious nor even original. Such a monopoly would moreover be economically undesirable. There is no good rationale for allowing anyone, not even automobile makers, to control the market of software publishing with property claims based on anything other than copyright.

It must suffice that the /user/ of a computer-aided automobile engineering invention needs to obtain a patent license, regardless of which software he uses. In practise, this usually does suffice. The only case where program claims would really perform an economic function is in the software industry, where the computer program as such would constitute the "invention"

The effect of allowing program claims is to make the publication of a program which can express the underlying invention a direct patent infringement -- regardless of how in fact the program would be used.

This would mean

* the program would be unusable for legitimate discussion and non-commercial experimentation, normally encouraged in patent law.

* the program would be unusable for legitimate purposes other than those specified in the patent application -- for example, a patented program method for predicting automotive engine dynamics could not be used to predict stock-market behaviour (and separate amendments saying that this is not infringement will not help, because such claims cover the distribution of said program and not only the use).

* EU companies would be forbidden to use the program method to compete against foreign companies overseas in territories where the patent was not in force.

Justification

Interoperability of data processing systems (e.g. computers) lies at the foundation of the information economy and allows for fair competition by all players large and small.

Article 6 of the Council only refers to the exemption provided for by the Copyright directive. This means that a software developer is allowed to find out how to make his data processing system interoperable with that of a competitor, but afterwards he cannot necessarily use his gained knowledge, since that could be covered by patents.

This amendment makes sure that patents also cannot be used to prevent interoperability. It was passed in an almost identical form by ITRE and JURI prior to the first reading ("data processing systems" read "computer systems or networks"). In first reading, a more sweeping version of this amendment was passed (with 393 vs 35 votes), which appeared as Article 9 in the consolidated version.

The expression "for the sole purpose" reverts to the spirit of the original ITRE/JURI version of the interoperability exemption (which is more limited), which was also supported by Luxembourg and several others in the Council (but didn�t make it).

Justification

This replacement is to be performed at all places in the text where the expression "computer-implemented invention" is used.

The expression "computer-implemented" is not suitable, because it implies that an invention can be wholly realised by means of a computer, which would mean that pure software is patentable. Since both the Commission and the Council agreed that software should not be patentable, the terminology used in the directive should not imply the contrary. The scope of the directive is thus the one of the patenting of devices that use software in order to aid the performance of the claimed invention.

The concept of a computer-implemented invention is not used by computer experts either, and in fact is not in wide use at all. It was introduced in May 2000 by the European Patent Office (EPO) to justify the patenting of "computer-implemented business methods" and bring EPO practice into line with Japanese and US practice. The term "computer-implemented invention" implies that solutions involving only generic computers are patentable inventions.

This idea is contrary to Article 52 of the European Patent Convention, which states that algorithms, methods for doing business, and computer programs do not constitute inventions within the meaning of patent law.The directive can not be intended to declare computer programs to be patentable inventions by presenting them in some other wording.

Justification

This amendment reflects current case law in Germany, and a similar decision in the UK case 'Gale's Application'.

In the words of the justices of the German Federal Patent Court (BPatG, decision of 26. March 2002, 17 W (pat) 69/98: