skip to main content skip to accessibility policy

Patents Newsletter

No. 4

April 2008

Article 4: The Return of the Computer Program Claim


David Alcock

In recent newsletters we have written about the extensive changes which have been proceeding in Europe - EPC 2000 came into force in December and The London Agreement has finally been ratified.

This edition focuses on further changes at the European Patent Office (EPO). This time they are less substantive but no less important to applicants as they relate to costs. 1 April 2008 saw the EPO increase the number of claims one may include in an application for ‘free’ to 15. However, the fee for any excess claims has substantially increased and are set to increase again from 1 April 2009. Increases are of such that the excess fees for a set of 60 claims rose by €6,750 from 1 April 2008. From 1 April 2009 the same set of claims will incur fees of €12,000! For applicants who continue with large claims sets it seems that the EPO would like their share of savings brought about by the London Agreement.

Back to top

Article 1: New EPO Claims Fees - What is the Effect?

As 1 April 2008 has passed, the EPO now charges €200 per claim on filing a European Patent (EP) application or on entry to the EP regional phase for each claim in addition to the 15th. The new claims fees can significantly increase the cost of filing an EP application having a larger claim set. Further increases will take effect for EP applications filed or entering the EP regional phase on or after 1 April 2009, when each claim in addition to the 50th will then cost €500. The increased claims fees are also payable at grant where the number of claims has increased since filing and there are more than 15 claims.

It will be some time before we see the full effect of the increased claims fees. Not only is there the direct effect on the pockets of applicants, there are potentially more subtle effects. New EP applications are likely to have fewer claims. As examiners become used to seeing smaller sets of claims, it is likely that they are going to be more reluctant to conduct a full examination of larger claims sets. Applicants are already seeing more objections under Rule 43(2) EPC where there is more than one independent claim per category and lack of unity objections, both a priori (based on independent claims) and a posteriori (based on dependent claims). Examiners are now going to be even more likely to raise such objections based on larger claim sets, rather than conducting a full substantive examination.

So how should we react to the increase in claims fees? Well, it is now even more cost effective to invest time on claim selection prior to filing to ensure that each claim really brings value to a claim set. The aim will be to provide a claim set with as close to 15 claims as possible, bearing in mind that each claim in addition thereto will have to justify its €200 price tag.

It will be even more important than ever to choose claims that have commercial value and to select dependent claims that provide a real fall back position.

A pragmatic approach to determining whether a claim is likely to be effective is to carry out a European-style problem and solution analysis, not just for the independent claims, but also for the dependent claims. This should help, not only to draft independent claims that only include the features essential to define the invention, but also to provide a set of focussed dependent claims.

Of course, the actual claim set for any particular application will need to take into account the nature and commercial importance of the invention. However, here are some things to think about:

  • In view of Rule 43(2) EPC2, which sets out that only in special circumstances may a European application contain more than one independent claim per category (product, process, apparatus or use), it may be appropriate to consider electing a single independent claim per category.
  • It may be worth considering electing a full set of dependent claims for the most important category only and/or to cross reference claim categories, (e.g. “A program product for carrying out the method of any preceding claim”). Taking into account the definition of a dependent claim in Rule 43(4) EPC2, it may be possible to build up a set of claims within a category using multiple claim dependencies to give maximum coverage, (e.g. “A system comprising the device of any preceding claim and...”). In this regard, it is to be noted that the EPO claims fees are computed based on the actual number of claims, irrespective of the dependencies.

Using such an approach may help to focus on the most important claims and to produce a compact set of claims. However, there may be other possible claims that one would not wish to discard completely. One approach is to include statements, for example at the end of the description, summarising in claim type language aspects that have been described in the application. As long as it is clear that such statements form part of the description, they are not counted as claims. Rule 137(4) EPC2 sets out that amended claims may not relate to unsearched subject matter. Accordingly, it may not be possible to subsequently move such statements into the claims set in the application in question. Also, it is likely that the statements will have to be deleted before grant, especially if they are inconsistent with the claims as allowed. However, the presence of such statements could give a clear basis for the filing of a divisional application during prosecution, if this becomes commercially of interest.

For further advice as to how to react to the increased claims fees, please contact your usual correspondent at D Young & Co.

Back to top

Article 2: Added Subject Matter

We are all aware of the pit-falls of added subject-matter. As the significance of a particular embodiment starts to increase, or an unknown document rears its head from obscurity, our ability to amend the claims using limitations present in the specification is closely scrutinised by Examiners, particularly those at the EPO.

As a result, it is absolutely critical that the application is drafted in such a manner which allows the applicant to make the necessary amendments as prosecution unfolds; however, recent experience has necessitated a closer look at this issue.

Increasingly, and worryingly, added subject-matter objections from the EPO seem to be focussing more than ever on combining preferred aspects of the invention and, in some cases, even combining dependent claims (yes, dependent claims!) Typical Boards of Appeal decisions on this issue, such as T0770/90 and T0296/96, support the viewpoint that the content of an application must not be considered to be a “reservoir” from which features pertaining to separate embodiments could be combined in order to create a particular embodiment. Further to this general principle, T0296/96 sets the high burden that any combination must have been “seriously contemplated” by the person skilled in the art. Whilst a number of subsequent decisions appear to have qualified this high burden to only require that the person skilled in the art should “consider” making such a combination (T1206/01 and T0036/06 to name but two), it appears that the EPO (at least in some quarters of chemical Examining Divisions) are taking a harder line than ever on this issue.

In some “extreme” cases, claims which consisted of a simple combination of multiply dependent claims have been rejected by an Examining Division for apparently violating Article 123(2) EPC. Although such a decision can always be appealed, this of course results in additional delay and expense to applicants.

What impact does this have on drafting? Well, if this attitude towards making amendments continues, it appears that applications will have to be drafted to explicitly disclose many more possible combinations and permutations of features so as to avoid “making a combination not explicitly or implicitly disclosed to the person skilled in the art at the filing date of the application”. Thus, in addition to including the most preferred features, it is advisable to consider at the outset those combinations which are most preferred. These highly preferred combinations can then be incorporated in the initial draft in preparation for any such onerous added subject matter objections which may arise later on.

This legitimate expectation that preferred features can be combined has even been approved by the Boards of Appeal, when it was stated that preferred combinations would be considered by the person skilled in the art, as these are “obviously the best way of achieving the technical features that the invention aims to provide” (T0068/99).

In summary, if this trend continues, applications will have to be drafted, at least from a European perspective, with considerably more specific combinations explicitly disclosed in the application as filed. Indeed, and as mentioned earlier, as prosecution unfolds and unknown prior art emerges, surely the only way to make sure that an amendment can be made without being scuppered by this harsh interpretation is to ensure that those preferred combinations are specifically included at the outset, instead of having to rely on being allowed to make such a combination at a later stage.

Back to top

Article 3: The London Agreement - Latest News

We reported in our February 2008 newsletter that the London Agreement will enter into force on 1 May 2008. The Agreement will apply to European patents granting after this date.

The London Agreement is aimed at reducing the cost of validating European patents by relaxing the translation requirements for validation in States that have signed up to the Agreement.

Previously it had not been certain whether the London Agreement would apply in Denmark or Sweden. This is because these countries had amended their national laws to take account of the Agreement, but had not formally confirmed their ratification of the Agreement with the German government (who are responsible for administering the Agreement). However, Denmark has now formally ratified the Agreement and it will apply in Denmark from its commencement date. Sweden has still not done this, but we are advised by our Swedish associates that their government has ratify the Agreement in time for its commencement.

Thus, we now know the London Agreement will apply in France, Germany, the UK, the Netherlands, Denmark, Switzerland, Liechtenstein, Luxembourg, Slovenia, Iceland, Latvia, Croatia and Monaco, and we expect it to apply in Sweden also.

The implementation of the Agreement in each State is governed by national law. This has led to some slight differences between the States. For example, in the UK the relaxed translation provisions already apply to European patents granted since 1 February 2008 – i.e. three months in advance of the official commencement date (of course, this is only relevant for non-English-language patents). The relaxed provisions have applied in some of the other States for even longer.

One potentially significant issue about the way the different countries have implemented the Agreement in their national laws concerns Germany. The German government seem to have made a mistake in the way they changed their national law back in 2003. As a result, according to the current German law, the London Agreement will only apply to patents granting after 1 June 2008! This is later than the Agreement requires and it is hoped the German government will take corrective action.

For previous articles regarding the London Agreement, please see our December 2007 and February 2008 newsletters.

Back to top

Article 4: The Return of the Computer Program Claim

It is well-known that the European Patent Convention (EPC) regards computer programs as excluded subject-matter, although this exclusion only applies to the extent that the relevant subject-matter is a computer program “as such”. UK patent law mirrors the provisions of the EPC.

According to case law from the European Patent Office (EPO), inventions relating to a technical process, such as image processing, are not treated as falling within the computer program exclusion, even if the invention is implemented using a computer program. It is therefore possible to obtain patent protection for such inventions.

Originally it was only possible to obtain system (computer plus program) and method claims in respect of these computer-implemented inventions. The trouble with system and method claims however is that they are generally infringed by the end user of a computer program, rather than directly by the supplier of the computer program. This end user is frequently a customer or potential customer of the patentee and therefore an unattractive target for patent litigation.

It is normally possible to attack a competitor who supplies a computer program used in an infringing system for indirect infringement. However, this approach is not available if the (potentially) infringing program is being supplied out of the jurisdiction. Furthermore, having to rely upon indirect (rather than direct) infringement is generally considered as increasing the uncertainty and risk for the patentee.

EPO practice changed following the IBM Decision in T1173/97, after which it became possible to claim a computer program directly. Such program claims are frequently expressed by reference to a method claim. For example, if claim 1 is an (allowable) independent method claim, then the computer program claim might be expressed as: “A computer program for implementing the method of claim 1”. With this terminology, it can be seen that the computer program claim is directed at essentially the same invention as the method claim, so the Board of Appeal held that it would not make sense if one were excluded and the other not.

In 1999, the UK Patent Office issued a practice notice aligning the treatment of computer program claims in the UK with the EPO IBM decision. However, the UK Patent Office had a change of mind after the Macrossan decision from the Court of Appeal in 2006. The Macrossan judgement did not specifically address the question of computer program claims, but did introduce a new four-step test for assessing excluded subject matter. This test included construction of the claim and an analysis of the contribution of the (supposed) invention. The view of the UK Patent Office was that if the claim were explicitly directed to a computer program, the contribution had to be construed as a computer program, and therefore fell into the excluded subject-matter exclusion. Accordingly, in a practice notice in November 2006, the UK Patent Office announced that following the Macrossan decision, it was no longer prepared to allow claims directed specifically to a computer program, even if the computer program implemented an allowable method.

This change in policy by the UK Patent Office was challenged by a group of companies, and this challenge was recently upheld by the High Court in the Astron Clinica decision. In this case, the judge (Kitchin J.) noted the desirability of maintaining alignment with the EPO, and also the problems that would be caused to patentees if computer program claims were not available. The judge further recognised that the change in practice by the UK Patent Office had not been explicitly mandated by the decision in Macrossan. Accordingly, the High Court overturned the newly adopted practice of the UK Patent Office.

In February 2008, the UK Patent Office (now rebranded as the UK Intellectual Property Office) issued a new practice notice indicating that it is not going to file an appeal against the Astron Clinica decision. Rather, the UK Intellectual Property Office will follow this judgement and allow computer program claims in cases where a corresponding method or system claim is allowable. In effect, this returns the situation in the UK to the pre-Macrossan position, and also brings back alignment with the EPO.

The decision in Astron Clinica to (re)allow computer program claims has generally been welcomed by practitioners, although not necessarily by the UK Intellectual Property Office, which described the ruling as “narrow”. Certainly it is true that the decision only impacts the type of claim that can be obtained, rather than more generally the type of invention that can be patented. Nevertheless, the judgement does indicate that the UK courts are prepared to accept and support the patenting of computer-implemented inventions in appropriate circumstances.

Back to top

Bookmark and Share

Related Resources

EPO Case G 3/08: Patentability of programs for computers

Related People

Partner, Patent Attorney
Partner, Patent Attorney
Partner, Patent Attorney
Partner, Patent Attorney

Follow us

Newsletter subscriptions

In support of our environmental policy we encourage email subscriptions to receive our patent and trade mark newsletters as soon as they are published. Please send your contact details to

For RSS users

Our RSS news feeds allow you to see when we have added new content to our website so you can get the latest site updates in one place, as soon as they are published.

Social media

Privacy Policy

We are committed to protecting and respecting your privacy. To understand our views and practices regarding your personal data and how we will treat it please see