Direct infringement requires making, using, selling, offering to sell, or importing an invention.1 It is often easy to identify makers, users, sellers, etc., for inventions defined by components or steps associated with a single location. For instance, all claimed components of a product invention might be amenable to inclusion in a single product.

Once an accused product is identified, it is normally easy to find a single person, company, or other legal entity that makes, uses, sells, offers to sell, or imports that product. For example, steps of a process invention might be carried out in a single location. If one can identify a single location where all claimed steps are performed, one can likely associate a single legal entity with activity at that location.

Claiming elements at a single location is frequently the most intuitive and logical way to define an invention, particularly if features of an invention will reside at a single location when that invention is actually practiced. Sometimes, however, a patent claim may define an invention by reference to components and/or activities at widely dispersed locations. In some such cases, the real innovation may reside in a combination of activities and/or components that will not normally be in a single place. Suppose, for example, that an inventor develops a technique for communicating large amounts of data across a network. Assume that technique requires pre-processing of data prior to transmission and post-processing the transmitted data once it is received. Further assume that there is prior art describing the same or similar preprocessing in an unrelated context, and that there is additional prior art describing a same or similar postprocessing technique in another unrelated environment.

Under those circumstances, it might be difficult to avoid the prior art by only claiming a transmitter or preprocessing steps or by only claiming a receiver or postprocessing steps. Even if there is a reasonable argument to distinguish such a claim over the prior art, however, there may be temptation to accept a multi-location claim. An examiner might indicate a willingness to allow a claim reciting steps at the transmitting and receiving ends, but might firmly reject claims that do not recite activity at both ends. Faced with the prospect of a lengthy and expensive appeal to obtain allowance of a single location claim, possibly combined with a limited patent budget and/or unknown market potential for the invention, the inventor or assignee may consider settling for the two-location claim.

Just as claims based on activities and/ or components at a single location will usually implicate a single legal entity, claims requiring components and/or activities at widely dispersed locations will often implicate multiple legal entities. This is sometimes referred to as “divided infringement,” as actions constituting infringement may be “divided” among different legal entities. Continuing the above example, an accused system may include a transmitter in one location and numerous widely distributed receivers. Company A may own and operate the transmitter, but the receiver may be owned and operated by individuals or other entities distinct from company A.


U.S. patent law regarding divided infringement is currently unsettled, particularly with regard to method claims. In a December 2010 opinion in the case of Akamai Technologies Inc. v. Limelight Networks Inc.2, a three judge panel of the Federal Circuit limited the circumstances under which a method claim could be infringed by the activities of multiple legal entities. That panel held that “there can only be joint infringement when there is an agency relationship between the parties who perform the method steps or when one party is contractually obligated to the other to perform the steps.”3 According to the panel opinion, there would not likely be an agency relation unless one joint infringer has the right to cause the other joint infringer to perform one or more claimed steps.4 If party A only has the right to control how party B performs a claim step, but party B is free to initially decide whether to perform or not to perform, there would likely be no agency sufficient to make A and B joint infringers. Similarly, the existence of a contract that gives party A the right to control how party B performs an activity would not create joint infringement if party B is not obligated to perform that activity. Stated differently, a contract under which party A can control the details of how party B performs a method step would not create joint infringement unless that contract also requires party B to perform that step.

The Federal Circuit recently vacated the December 2010 Akamai opinion and agreed to rehear the appeal en banc (i.e., before all judges of the Federal Circuit).5 The Federal Circuit specifically asked the parties to file new briefs to address the following issue: “If separate entities each perform separate steps of a method claim, under what circumstances would that claim be directly infringed and to what extent would each of the parties be liable?”6 The Federal Circuit has also agreed to en banc rehearing of the separate case of McKesson Technologies, Inc. v. Epic Systems Corporation and has asked the parties in that case to address the circumstances under which there may be liability for inducing infringement or for contributory infringement when separate entities perform method claim steps.7

In its principal brief, the Akamai appellant has argued that “[a] method claim is directly infringed when every step of the claim is practiced in the United States, whether by a single entity or by entities whose actions combine to perform all steps of the claim.”8 Briefs of numerous amici curiae have urged that separate entity infringement should be determined using standards that would be more encompassing than the agencyrelationship- or-contractual-obligation standard of the vacated Akamai opinion. As of the date this article was written, the Akamai appellee had not yet filed its brief.

This article takes no position on how the Akamai or McKesson appeals should be decided, but it seems reasonably possible that the Federal Circuit will at least partially retreat from the standard of the vacated Akamai opinion. As noted by the Akamai appellant and several amici, the Supreme Court has resisted Federal Circuit efforts to create bright line rules in patent law. Requiring an agency relationship or contractual obligation, in the absence of a statute clearly imposing such requirements, seems to be the type of standard the Supreme Court might reject.


In addition to method claims reciting steps in multiple locations, patents may also include device claims that recite components in multiple locations. Multi-location device claims, which are often styled as “system” claims, may recite elements that are owned and operated by different legal entities. For example, a claim might recite a server computer and a client computer. In practice, the server may be owned and operated by one entity and the client might be owned and operated by a different entity. In some cases, one of those entities might be an infringer.

In Centillion Data Systems LLC v. Qwest Communications International Inc.9, Federal Circuit held that “to ‘use’ a system for purposes of infringement, a party must put the invention into service, i.e., control the system as a whole and obtain benefit from it.”10 However, a party that uses a system under Section 271(a) need not exercise physical or direct control over each element of that system.11 Although a party must use every element of the system in order to infringe, that party can do so by placing all elements of the system collectively into service.12 Notably, one entity that puts a system into service can be an infringer even if another entity physically possesses other elements of the system.13

It is unclear how the Centillion rule might be affected by the en banc decisions in Akamai and McKesson. Although the Centillion rule is based on principles that differ (at least in part) from the principles at issue in Akamai, the Centillion opinion did cite to the nowvacated Akamai opinion for certain aspects.


The law of divided patent infringement will remain in flux while the Akamai and McKesson cases are pending. For the present, it is prudent not to rely solely on patent claims that require actions by multiple legal entities.