This article provides an overview of how divided patent infringement is viewed in Germany.  

The article is addressed to US Patent Counsel dealing with international patent portfolios. There are situations, inparticular in the computing arts context, where inventions involving methods or systems might be put in practice by a plurality of actors. Related strategic considerations from a German perspective are highlighted in the present article.   

A party that practices every element of a patent claim is a direct infringer.  There is essentially no difference between the United States and Germany regarding this definition.  However, a situation may arise in which no single party can be found to practice all elements of a claim, but multiple parties combine to do so.  This situation, known as “divided infringement” or “joint infringement,” may occur in cases of method claims, in which multiple actors individually perform the different method steps, and only their combined efforts cover all elements of the patent claim.  For example, if a claim has three steps, and one entity performs the first step and another entity performs the next two steps, each entity could assert that it is not liable for infringement because it does not perform all steps of the method.  The same issue may apply to system claims, in which the system is put in service by different people.  Of particular importance are claims in the computing arts context, involving, for instance, actions or functionalities to be implemented on a server side and a client side.  

The standards of divided infringement are in motion in the United States.  In BMC Resources¹, the US Court of Appeals for the Federal Circuit held that in cases of divided infringement of a method claim, an infringer must “control or direct” each step of the method for there to be direct infringement.  Later, in Centillion Data Sys.2, the Federal Circuit decided on a system claim and ruled that there is direct infringement if the infringer “puts the system as a whole into service, i.e., controls the system and obtains benefits from it.” On 18 November 2011, the Federal Circuit heard oral arguments in two separate appeals (Akamai and McKesson)³ to review the legal standards for divided infringement.  It thus remains unclear to what extent claim drafters must change their prosecution strategies in the future. 

How would a German court deal with issues of divided infringement?  In a smaller market such as Germany, this question might even have a territorial aspect.  The actors may be from different countries—for example, the server may be hosted in the United States while the client’s computer is located in Germany.  The concept of territoriality requires that the infringement take place in Germany; otherwise there is no infringement.   

The actor may directly or indirectly infringe the patent.  That is, even if there is no direct infringement, the actor may still be liable for actively inducing patent infringement or for contributing to infringement (so-called indirect infringement).  Although the US and German patent systems both provide for these categories, the concepts are not the same.     

Indirect Infringement – the German Concept   

The concept of indirect infringement is provided in Sec. 10 of the German Patent Act: “A patent shall have the further effect that a person not having the consent of the patentee shall be prohibited from supplying or offering to supply within the territory to which this Law applies a person, other than a person entitled to exploit the patented invention, with means relating to an essential element of such invention for exploiting the invention, where such person knows or it is obvious from the circumstances that such means are suitable and intended for exploiting the invention.” 

Thus, indirect infringement requires a purpose-oriented supply of a “means relating to an essential element of the invention”:   

Purpose-oriented supply.  Sec. 10 does not require that the contributory infringement finally lead to a direct infringement.  The supplied means must only be suitable for putting the invention into effect and be intended for that purpose.  Thus, liability for this type of infringement requires a certain level of knowledge on the side of the contributor.  The patentee is only required to demonstrate that—in consideration of the circumstances—the supplied means is obviously suitable and intended for an infringing use.  Of course, if the supplied means can only be used in an infringing manner (i.e., it does not allow any non-infringing use), that knowledge will be assumed by the courts.

Supplied means. The supplied means must at least relate “to an essential element of the invention” that is capable of interacting functionally with one or more features of the patent claim to implement the protected invention.  An element that is completely subordinate to the technical teaching of the invention would be a non-essential element and thus not sufficient for a contributory infringement.  Even means that are staple commercial products can be means suitable for contributory infringement.  The supply of staple commercial products would, however, require the deliberate inducement of the purchaser to commit a direct infringement.  A person who actively and intentionally induces or aids a third party to willfully commit an infringement of a patent is also liable. 

The concept of indirect infringement would cover most cases of divided infringements.  In contrast to the US model, the German concept provides significant room to find infringement.  Nevertheless, the contributor can only be prohibited from supplying and offering means within the territory of Germany.  Additionally, the legal consequences of an indirect infringement are more limited than those of direct infringement.  A patentee should rely on direct infringement rather than the more problematic case of indirect infringement.  Accordingly, it is important to know where the borderline between direct and indirect infringement is situated. 

Borderline Between Direct and Indirect Infringement

In a decision rendered in October 2009 (known under the catchphrase “MP3-Player Import”), The Federal Supreme Court⁴  gave at least some guidance on how to find this borderline.  The court ruled that “an injunction- and destruction claim may not only be asserted against someone, who in his own person implements one of the infringing acts defined in paragraph 9 Patent Act [=direct infringement] or who intentionally enables or facilitates the implementation of the infringing act by a third party.  Direct infringer is also who enables or facilitates the implementation of the infringing act by a third party, although he could obtain with reasonable effort knowledge that the act supported by him infringes the absolute right of the patentee”. 

Germany’s Federal Supreme Court apparently establishes low thresholds for direct infringement requiring only that the infringer supports a third party.  These standards would thus not require that the infringer “controls or directs” the steps of the method or that it “puts the system as a whole into service, i.e., controls the system and obtains benefits from it”, as it would be required by the present US case-law.  The court’s ruling, however, must be concretised in the particular context of divided infringements.  

Divided Infringement of Method Claims 

The above-mentioned ruling of the Federal Supreme Court may not mean that preparatory acts alone would be sufficient for a direct infringement.  On the other hand, it is not required that the actor necessarily implement all steps of a method claim to be guilty of infringement.  The actor would still directly infringe the patent if, for example, the final steps of the method were implemented by a third party.  This has already been ruled in earlier decisions of the District Court Düsseldorf⁵, which is one of the three prominent patent infringement courts in Germany.  

A recent decision of the Appeal Court Düsseldorf (known under the catchphrase “Prepaid Cards”) further substantiates the approach⁶.  The case concerned a situation where one half of the method steps were committed in Germany, and the other half outside.  Patent infringement requires acts occurring in German territory, whereas exterritorial activities may not have the same effect.  A claim of a German patent (or the German part of a European patent) cannot be infringed by acts occurring outside the German territory.  Therefore, the decision contains a general statement on how to deal with divided infringements, because it equally measures the minimum requirements that must be fulfilled by someone (in Germany) necessary for a conclusion of direct infringement of the patent.    

Subject to that decision was a patent asserted against so-called “Prepaid Cards”.  It covered a method of processing prepaid telephone calls with the following language:  

“A method of processing telephone calls, particularly for use in connection with public telephones, comprising  

(a) programming a respective Public Automatic Branch exchange (PABX) to become toll-free accessible for incoming calls through dialing any one out of a series of predetermined numbers stored in a data-bank of the PABX; 

(b) enabling a calling party to complete a connection with a called party; 

(c) cutting-off the said connection after a prefixed time/counter pulses interval; 

(d) erasing from the data-bank any number that had once been dialed; 

(e) marking the said series of numbers, each on a vendible  carrier member in an invisible—however readily exposable—manner; and 

(f) offering the vendible carrier members for sale to the general public.” 

The idea behind this method is that someone buys a “prepaid card” (“vendible carrier members”) with a dial-in number and an imprinted PIN hidden under a scratch-off layer on the card (“invisible—however readily exposable—manner”).  By dialing the dial-in number and then the PIN, for instance from a public phone, the caller reaches the computing system of the service provider.  The computing system is programmed to enable the caller to connect with someone elsein compliance with a predetermined credit.  Upon entering the telephone number, the computer system establishes the connection between the parties.  Once the predetermined credit is used up, the computer system cuts off the connection.  

Some of these steps were realised in Germany, whereas other steps were realised outside Germany.  The programming of the computing system, comprising a data bank in which the PINs of the calling cards were stored and with which a connection was established, depending on the remaining credit attributed to a certain PIN and the corresponding calling card, did not take place in Germany.  Nevertheless, the  Appeal Court Düsseldorf concluded that there was direct infringement.  The reasoning of the court can be summarised as follows:  

• Direct infringement of a method claim does not require that necessarily all method steps are implemented in German territory. 
• The steps committed outside German territory must be imputable to the person implementing the steps within German territory.  “Acts implemented abroad are then treated like acts implemented in Germany”.  
• Imputability requires a purpose-oriented activity of the actor such that the advantages of the claimed invention take effect in Germany. 

This reasoning might be mirrored in the normal situation of a divided infringement.  An entity performing only one step (with another entity performing the remaining steps) of a method claim could assert that it is not liable for direct infringement because it does not perform all steps of the method.  As long as the missing steps can be imputed to that entity, however, it is a direct infringer.   

Divided Infringement of System Claims  

System claims are treated like product claims.  According to Sec. 9 of the German Patent Act, “a person not having the consent of the patentee shall be prohibited from manufacturing, offering, putting on the market or using a product which is the subject matter of the patent or importing or stocking the product for such purposes”.  At least one of the indicated acts must be fulfilled in order for there to be infringement.  

The question of divided infringement may arise if one entity supplies only one element of the claimed system and another entity combines it with other elements to the complete system.  The question would then be whether the “offer” of that element represented an offer of the whole system, or whether the manufacture of the element could be considered as “manufacturing” according to Sec. 9 of the German Patent Act.  

The answer to that question requires a case-by-case analysis.  However, direct infringement is likely if the supplied element represents an essential component of the claimed system which is supposed to be completed by other elements of the system.⁷  The delivery of the element would likely be considered “putting on the market”, in the sense of Sec. 9 the steps of  of the German Patent Act, even though another entity adds missing pieces to complete the system.  If the element is also produced by the entity, its production would consequently fulfill the requirement of a “manufacturing” according to Sec. 9 of the German Patent Act.  

The approach developed by the Appeal Court Düsseldorf8 for method claims gives guidance on how to consider the infringement of system claims, in particular if the “use” according to Sec. 9 is in question.  The use of the system may be carried out by two or more actors, who may even be situated in different parts of the world.  For instance, a server may be hosted in Germany, while the clients are situated in other parts of the world, or vice versa.  

The reasoning of the Appeal Court Düsseldorf appears to be equally applicable to the use of a system; it doesn’t appear to be limited to the use of a method.  Both uses relate to an activity.  Accordingly, it is likely that a German court would conclude that an infringing use of a system would not require that the used components of the system be located within Germany, as long as the acts committed outside German territory can be imputed to someone acting in Germany.  This imputability would then require a purpose-oriented activity of the actor such that the advantages of the claimed invention take effect in Germany.  

Translating these assumptions to a normal situation of divided infringement where the infringement is simply divided between different actors (each acting in the German market), a German court would likely consider the actor that enables or facilitates the use of the system by other actors to be the direct infringer.  

Even though German case law gives only dim guidance on where the border between direct and indirect infringement of system claims lies, German courts are certainly more flexible than US courts in concluding direct infringement.  German courts would not require that a direct infringer puts the system as a whole into service, i.e., controls the system  and obtains benefits from it, as the Federal Circuit stated in Centillion Data Sys.9  The German thresholds are lower.  Of course, in the context of computing arts in particular, it is certainly appreciated to have claims which consider the system from all possible angles.  In a server-client situation, it might be useful to consider the invention from the back end to the front end.  Doing so may later give hints to the courts to indentify the benefits of the invention and render it easier to identify a direct infringer.

_{1 BMC Resources, Inc. v. Paymentech, L.P., 498 F.3d 1373 (Fed. Cir. 2007).}

_{2 Centillion Data Sys. v. Qwest Communs. Int'l, Inc., 631 F.3d 1279 (Fed. Cir. 2011).}

_{3 Akamai Technologies, Inc. v. Limelight Networks, Inc. and McKesson Technologies, Inc. v. Epic Systems, Corp.}

_{4 BGH, Xa ZR 2/08.}

_{5 District Court Düsseldorf, GRUR RR 2001, 201, 204, InstGE 1, 26, 31g.}

_{6 Appeal Court Düsseldorf, 2 U 51/08.}

_{7 Federal Supreme Court, BlPMZ 51, 322, Mülltonne.}