Induced infringement in the United States expanded to include divided infringement

Induced infringement in the United States expanded to include divided infringement

Induced infringement in the United States expanded to include divided infringement

Akamai Technologies, Inc. v. Limelight Networks, Inc. and McKesson Technologies, Inc. v. Epic Systems Corp. (Fed. Cir. 2012)

In September 2012, the full bench of the United States Court of Appeals for the Federal Circuit handed down a decision that dramatically alters the previous understanding of the law of induced infringement, strengthening the law for patentees in the process. In Akamai, the Federal Circuit held that induced infringement can be found where multiple entities collectively perform all of the steps of a patented method.

Infringement law prior to the Akamai decision

Direct infringement is covered by 35 U.S.C. § 271(a). The courts have long held that for direct infringement to occur, a single actor and/or an agent of that actor (e.g. a contractor under the control of the actor), must practise each and every element of the claimed invention. This is known as the “single-entity rule”.

Induced infringement is covered by 35 U.S.C. § 271(b), which states that “[w]hoever actively induces infringement of a patent shall be liable as an infringer”. Prior to the Akamai decision, the courts had interpreted induced infringement as requiring that:

  1. the alleged infringer knowingly induced infringement and possessed specific intent to encourage another’s infringement1; and
  2. the inducement leads to direct infringement by a single actor.2

Thus, both direct and induced infringement required infringement of all steps of a method by a single actor. As a result, divided infringement was often asserted as a defence in instances where a party did not perform all steps of a method claim.

Background to the Akamai v Limelight decision

Akamai is the owner of US Patent No. 6,108,703, directed to methods for the efficient delivery of web content. Akamai accused Limelight of performing all but one of the steps of Akamai’s patented method of delivering web content, with Limelight’s customers performing the last step, pursuant to Limelight’s instructions. Akamai asserted that Limelight had sufficient control over its customers to be considered an agent relationship, and should therefore incur liability for joint direct infringement. In 2010, a Federal Circuit panel held that there was no joint direct infringement because Akamai failed to show an agency relationship and thus failed to demonstrate that a single actor performed all of the steps of the method.3

McKesson owns US Patent No. 6,757,898 directed to methods for electronic communication between healthcare providers and their patients. McKesson alleged that Epic Systems Corp. induced Epic’s customers to infringe the patent. Epic did not perform any steps of the claimed method. Rather, some steps were performed by the patients and some were performed by the healthcare providers. In 2011, a Federal Circuit panel held that there was no infringement because Epic’s direct customer, the healthcare providers, did not perform all of the steps of the method.4

The U.S. Court of Appeals for the Federal Circuit agreed to a rehearing en banc of these decisions to consider if, and under what circumstances, there can be joint liability when at least two parties collectively perform all of the steps of a patented method, but no one party performs all of the steps.

The decision to expand the definition of induced infringement

In a close 6-5 decision, the court overturned precedent and ruled that induced infringement under 35 U.S.C. §271(b) does not require a single, direct infringer (as previously held), but can involve two or more parties that collectively perform all of the steps of a patented method. In arriving at this decision, the majority observed that a party who knowingly induces others to engage in acts that collectively perform the steps of a patented method should not be immune from liability simply because they have delegated infringement activity.

This decision expands the definition of induced infringement to include divided infringement. The court re-iterated that the element of intent to infringe must be shown to establish induced infringement. Although there remain no clear guidelines as to what is required to prove intent, at the very least the inducing infringer should be shown to have had knowledge of the asserted patent.

With this new understanding of induced infringement, inducement liability exists where the accused infringer:

  1. was aware of the patent and the patented method;
  2. induced one or more parties to perform the steps of the patented method; and
  3. those steps were performed, either by a single induced party, collectively by multiple induced parties or collectively by the accused infringer and one or more induced parties.

The majority’s decision was met with strong dissent from the other five members of the court. Both defendants, Limelight and Epic, subsequently petitioned the Supreme Court to review the induced infringement question, although Epic has now withdrawn their petition following a settlement with McKesson. Limelight’s petition is still pending.

Implications of the decision

Decision strengthens the law of induced infringement for patentees

This decision strengthens the law of induced infringement for patentees, with divided infringement no longer a defence for accused infringers. In light of the expanded definition of induced infringement, those with patented methods in the United States should review the practices of competitors in that country to determine whether induced infringement may be occurring. Those with business in the United States should also review their own practices to ensure that they are not liable for infringement under these new rules.

Claim drafting

The decision also has implications for claim drafting. Where possible, method claims were previously drafted to ensure that a single entity would perform all of the steps. While this is still preferable, there is now more flexibility to include claims that contain steps that may be performed by multiple parties, as such claims now have added value.

Implications for the Biotechnology industry

The expanded definition of induced infringement under Akamai has particular implications for the biotechnology industry and the information and communications technology (ICT) industry, both of which commonly utilise methods that may require multiple entities to perform the steps.

Diagnostic and personalised medicine methods may be particularly affected by the Akamai decision. These methods often include steps carried out by multiple users or entities, such as laboratories, pharmaceutical companies, technicians, nurses, doctors and even patients. Patented methods that require multiple users may now be more easily enforceable. Of course, where possible claims to these types of methods were drafted to ensure that a single entity, such as a laboratory technician, would perform all of the steps. While this practice should certainly continue, additional emphasis can now be given to claims that include steps performed by other entities.

The expanded definition of induced infringement can be used to strengthen personalised medicine method claims. Some of these types of claims have recently come under attack as not being patent-eligible subject matter under 35 USC § 101 in the wake of Mayo v Prometheus. Claims to personalised medicine methods, such as those that were held invalid in Mayo v Prometheus, could be drafted to include, for example, a treatment step to help satisfy the 35 USC § 101 requirement. Such claims would typically require at least two parties to perform the method (e.g. a laboratory or clinical technician and a nurse or doctor), but would be enforceable against induced infringement under the new standards set by Akamai.

Implications for the ICT industry

Inventions within the ICT industry have long been recognised as having the potential to include methods involving multiple entities. This is not only exemplified in the patents considered in the Akamai decision, but also in the majority’s decision itself. In discussing the legislative history of contributory infringement, the majority point to a 1948 hearing in which the speaker conceives of a new radio communication method necessarily including both transmitting and receiving steps.5 In more recent times, growth in fields such as distributed computing and cloud based applications, only further highlight this issue.

With this in mind, whilst claims directed towards ICT methods should continue to be drafted such that all steps are performed by a single entity, claims including multiple entity steps may also be considered for those seeking protection in light of Akamai, for example, to provide additional fallback positions. However, this approach should be used with caution, particularly in respect of cloud-based inventions and other distributed ICT methods which have the potential for multiple-entity steps to also be performed in different jurisdictions. Attempting to assert patented method rights against multiple entities across multiple jurisdictions, for example in cloud-based applications utilising off-shore servers, may introduce additional complication.


  1. DSU Med. Copr v JMS Co., 471 F.3d 1293, 1306 (Fed. Cir. 2006) (en banc)
  2. BMC Resources, Inc. v. Paymentech, L.P., 498 F.3d 1373 (Fed. Cir. 2007)
  3. Akamai Technologies, Inc. v. Limelight Networks, Inc., 629 F.3d 1311 (Fed. Cir. 2010)
  4. McKesson Techs., Inc. v. Epic Systems Corp., No. 2010-1291, 2011 WL 1365548 (Fed. Cir. Apr. 12, 2011)
  5. Akamai, p 19.