A finding of liability for direct patent infringement in the U.S. is typically based on the infringing actions of a single entity, either a natural or legal person.

However, under the doctrine of “divided infringement” (sometimes referred to as “joint infringement”), a single entity can perform fewer than every step of a patented method or make use of fewer than every element of a patented system and still be liable for direct infringement.

Such liability for divided direct infringement is separate from liability for indirect infringement—under which an entity can be liable for contributing to or inducing the direct infringement of other entities—since liability for indirect infringement can be imposed only once direct infringement, whether based on the actions of a single or multiple entities, has been proven. (Limelight Networks, Inc. v. Akamai Techs., Inc., 134 S.Ct. 2111, 2117-18 (2014)).

All else being equal, divided direct infringement is more likely for networked inventions, for which the claims can easily encompass the actions of different entities, and for those high-tech and biopharmaceutical inventions that incorporate judicially created exceptions to patent eligible subject matter (“abstract ideas”, “laws of nature” and “natural phenomena”), for which the claims, under the Alice/Mayo framework, must include additional patent eligible subject matter. For example, an otherwise patent ineligible diagnostic method can be rendered patent eligible in the U.S. by inclusion of a treatment step, opening the possibility for divided infringement based on separate entities performing diagnosis and treatment. (MPEP § 2106.04(d)(2) (9th Ed., Rev. 07.2022) citing Vanda Pharm. v. West-Ward Pharm., 887 F. 3d 1117, 1135-36 (Fed. Cir. 2018)).

A single entity that performs only some of the steps of a patented method can be liable for direct infringement if (1) the single entity “directs or controls” performance of the remaining step or steps by the other entities, by conditioning their participation in an activity or receipt of a benefit upon their performance and establishing the manner or timing of that performance, or (2) if all entities form a joint enterprise. (Akamai Techs., Inc. v. Limelight Networks, Inc., 797 F.3d 1020, 1022 (Fed. Cir. 2015) (en banc)).

A single entity that makes use of only some of the elements of a patented system can be liable for direct infringement if the single entity “control[s] the system as a whole and obtain[s] benefit from it”. (Centillion Data Sys., LLC, v. Qwest Comms., 631 F.3d 1279, 1284 (Fed. Cir. 2011) quoting NTP, Inc. v. RIM, Ltd., 418 F.3d 1282, 1317 (Fed. Cir. 2005)).

The divided infringement inquiry for method or system claims thus involves not only mapping a set of allegedly infringing actions onto claim limitations, but also requires considering exactly which entities are responsible for exactly which actions and how those entities relate to each other and to the actions performed by the other entities. Such an inquiry is necessarily highly fact- and context-dependent and can involve considerations that include the contractual, regulatory or customary obligations of the different entities and the extent to which one entity provides technical support, instructions or guidance, or sells or loans materials or equipment to other entities. The issue of divided infringement has been disputed in a range of contexts, from doctors instructing patients to ingest a particular vitamin before receiving anti-cancer medication (direct infringement) to providers of portable credit card readers and software to merchants for use in processing customer payments (possible direct infringement).

For patent owners, method or system claims that could require the collective action of multiple entities for infringement thus risk being difficult or impossible to enforce and are best avoided, whenever possible. In the telecommunications sector, the divided infringement doctrine has been largely circumvented by adopting claims that cover less than an entire system or describe a method from the point of view of a particular element, e.g., a mobile device, a server device, etc. However, in technologies that encompass patent ineligible subject matter, and are therefore constrained under the Alice/Mayo framework to claim additional patent eligible subject matter, avoidance may not be an option, making it necessary to consider, for a given method or system claim, how the doctrine of divided infringement would be applied to the potentially infringing collective action of multiple entities.

Conversely, for potential infringers, method or system claims that could require the collective action of multiple entities to be infringed present opportunities for adaptation—for example, by adjusting how the potentially infringing actions are distributed between the multiple entities or how the multiple entities relate to each other or to the potentially infringing actions performed by other entities—that can limit or even eliminate liability for divided infringement.

In sum, when presented with U.S. method and system claims that could require the collective action of multiple entities to be infringed, it is necessary to consider the highly fact- and context-specific divided infringement doctrine.