By: Audrey A. Millemann

In Akamai Technologies, Inc. v. Limelight Networks, Inc., 692 F.3d 1301 (Fed. Cir. 2012), an en banc Federal Circuit Court of Appeals was divided over the issue of “divided infringement” in the context of inducing infringement. 

A party is liable for inducing infringement if it instructs or causes another party to infringe a patent.  Thus, there is an act of direct infringement that results from the inducement.  “Divided infringement” of a method patents exists when multiple parties perform different steps of the claimed method.  

Before Akamai, there could be no liability for inducing infringement unless a single entity was liable for direct infringement (the “single-entity rule”).  See, BMC Resources, Inc. v. Paymentech, L.P., 498 F.3d 1373 (Fed. Cir. 2007).  In this case, the Court of Appeals overruled BMC, holding that “all the steps of a claimed method must be performed in order to find induced infringement, but that it is not necessary to prove but that all the steps were committed by a single entity.”  Thus, under the court’s holding, there is no requirement that a single party must be liable as a direct infringer.

The Akamai case consisted of a rehearing en banc of two separate cases in which the plaintiffs sued the defendants for inducing infringement.  The first case, Akamai v. Limelight, involved a patent for a method of delivering website content.  The defendant performed some of the steps of the claimed method and instructed others (its customers) to perform the remaining steps.  The district court for the District of Massachusetts granted judgment as a matter of law for the defendant.  In McKesson Technologies, Inc. v. Epic Systems Corp., the patent covered a method of electronic communications between healthcare providers and their patients.  The defendant performed none of the steps of the claimed method, but instead instructed the providers and the patients to perform all of the steps.  The district court for the Northern District of Georgia granted summary judgment for the defendant. 

The court granted a combined rehearing in the two cases because the cases involve similar issues concerning inducing infringement of method patents.  The issue in Akamai was whether the defendant was liable for inducing infringement if the defendant performed some of the steps and induced other parties to perform the remaining steps.  The issue in McKesson was whether the defendant was liable for inducing infringement if the defendant induced multiple other parties to perform all of the steps and the defendant itself did not perform any of the steps. 

In its decision, the court discussed the issue of divided infringement in the context of direct infringement.  The court stated that it was not required to resolve the issue of whether a single party must perform all of the steps of a method for direct infringement to be found.  Direct infringement, as set forth in 35 U.S.C. section 271(a), is a strict liability tort.  To be liable, a defendant must perform all of the steps of the claimed method, either personally or vicariously (through another person under the defendant’s control, such as an agent).  There is no direct infringement if several independent persons perform the steps of the method, even if the independent parties specifically agree to divide up the steps of the method in order to avoid liability for patent infringement.  The court explained the rationale as follows: it would not be fair to hold independent (non-agency) actors liable because independent actors would have no way of knowing of each other’s actions or of knowing that together they infringed the patent.   

In contrast, the court explained, inducing infringement is not a strict liability tort.  The inducer must have knowledge that the acts that it induces constitute patent infringement.  Specific intent is required.  See, DSU Med. Corp. v. JMS Co., 471 F.3d 1293 (Fed. Cir. 2006).  Thus, because of the specific intent requirement, the risk of holding unknowing persons liable for inducing infringement is less likely than in a case in direct infringement. 

Under DSU, to be liable for inducing infringement under 35 U.S.C. §271(b), a defendant must actively induce another party to commit infringement, by advice, encouragement, urging, or aiding the other party.  The other party who commits the infringement does not need to be the agent or under the control of the defendant, as is the case for a defendant to be vicariouisly liable for direct infringement.  To be liable for inducing infringement, however, there must be an act of direct infringement.  As the court explained:  “The reason for that rule is simple:  there is no such thing as attempted patent infringement, so if there is no infringement, there can be no indirect liability for infringement.”

In BMC, the court had held that the act of direct infringement had to be committed by one entity.  The court based its decision on two legal principles:  first, that inducing infringement requires an act of direct infringement; and, second, that direct infringement requires a single person to perform all of the steps in the claimed method.  The Akamai court explained that these legal principles are correct, but that the BMC court’s conclusion was wrong – inducing infringement does not require that a single party be liable for the act of direct infringement.  The court said that it would not be fair to allow an inducer to escape liability for inducing infringement just because the inducer instructed multiple parties to each perform different steps and no single party performed all of the steps.  There is no policy or legal basis for immunizing this inducer and holding liable an inducer who instructs one party to perform all of the steps of the method.  There is also no basis for immunizing an inducer who performs certain steps itself and instructs others to perform the remaining steps. 

The court discussed the legislative history of §271(b) and concluded that the drafters had intended to cover acts of infringement by persons acting in concert where none of the parties were separately liable for direct infringement.  The court also analyzed joint liability in criminal law and tort law, and found support for its position. 

Five judges dissented.  In their dissents, these judges argued that the majority’s decision was a “sweeping change” to existing patent law.  According to the dissent, subsection (b) is clear in its requirement that there be an act of direct infringement in which all of the steps are performed by a single entity.  These judges strongly argued that the majority was rewriting the statute and should defer to Congress for clarification. 

The majority responded that the relevant question was what Congress had intended in enacting the statute in 1952 and that the decision in Akamai was not new law, but the correct interpretation of existing law.  The court stated:

“[W]e are persuaded that Congress did not intend to create a regime in which parties could knowingly sidestep infringement liability simply by arranging to divide the steps of a method claim between them… it is unlikely that Congress intended to endorse the ‘single entity rule,’ at least for the purpose of induced infringement, advocated by [the defendants], which would permit ready evasion of valid method claims with no apparent countervailing methods.”


The Court of Appeals reversed both cases, holding that the defendants in each case could be found liable for inducing infringement if:  (1) the defendant knew of the plaintiff’s patent; (2) the defendant induced the performance of the steps of the claimed method (or the defendant performed one or more steps itself and induced others to perform the remaining steps); and (3) the steps of the claimed method were actually performed. 

This decision is sure to be viewed as a victory for patent owners and likely to lead to more complicated patent litigation involving multiple parties.