[Guest post] Closing the patent loophole across borders

Patents set themselves apart from other IP rights as inventions are often composed of multiple physical components or steps in a method, which does not necessarily have to exist or be performed at the same time and place. When an infringing act becomes "divided" across borders in this way, it raises the question where an infringement occurs, if at all.

Katfriend Mattias Rättzén (Taylor Wessing and Harvard LL.M.) has written about this in an article that was recently published in the UIC Review of Intellectual Property Law. In this guest post, he summarizes his main findings.

Here's what Mattias writes:

Closing the patent loophole across borders

by Mattias Rättzén

Extra-territorial Kats

It is a legal maxim that patents, like any other intellectual property rights, are territorial rights. Patents are granted on a national basis and so patent holders seeking protection in multiple countries are invited to procure multiple national rights. Historically, courts, and most notably in the United States, have opted for a narrow understanding of the concept of territoriality of patents. This has more recently been reaffirmed by the United States Supreme Court in Deepsouth Packing Co. v. Laitram Corp, 406 U.S. 518 (1972) and Microsoft Corp. v. AT&T Corp, 550 U.S. 437 (2007) where the Court refused to extend the U.S. Patent Act extraterritorially based on the presumption that federal statutes, including patent statutes, shall not extend to acts occurring abroad.

The traditional notion that patents are territorial rights is, however, under considerable stress. Just as commerce is becoming increasingly global, so do infringing activities coincide less neatly with national borders. Depending on the technical field, the subject-matter is also becoming increasingly prone to disseminate across borders. Because patent statutes differ between different jurisdictions, there is and will continue to be a tension between these economic and technological developments and the territorial limitation of patent rights.

What are we talking about? Let us take a simple hypothetical example. Assume that a patent covers a system for a machine learning model, which involves the use of computing devices configured to receive sourced data, process data, configure an algorithm and to complete a particular machine learning model. For each claimed configuration there is a set of routine operations to be completed, including receiving requests to perform the computational services and validating such requests. Further assume that a third party uses the same machine learning technique to create a mobile application that sources substantial amounts of user data and then reads and transforms that data into usable content in the application. If each operation is performed on servers in different countries, and if users whose data is sourced and users who use the mobile application reside in different countries, some where there is a patent and some where there is none, where does a prospective infringement take place? A number of possibilities could be thought of, ranging from where data is collected and where servers are located to where users are residing. Each point of contact could in theory be sufficient to render liability, but that seems to ignore that a full and complete infringement is not performed anywhere. The inclination to catch all infringing acts in one go clearly struggles with the fact that patent protection remains territorial.

The implications of this debate are far from hypothetical or theoretical. Inventions such as the one described are becoming increasingly commonplace, and are present in many other technical fields such as in biotechnology and telecommunications. In each of these areas, it is becoming increasingly difficult for both right holders and users to predict in which jurisdictions inventions are deemed to be carried out. There is currently no universal solution in place to articulate clear and uniform criteria for balancing long standing territorial legal norms with what are new technological and commercial realities.

So how have courts tackled these issues so far? Some courts in the United States have persisted, even if inconsistently, in refusing to find for infringement when infringing activities that fall under patent claims have been geographically divided. In NTP, Inc. v. Research in Motion, Ltd., 418 F.3d 1282 (Fed. Cir. 2005), the Federal Circuit made a distinction between system claims and method claims, and found that there was no infringement of the latter where not all steps were performed domestically. This outcome is akin to the approach taken earlier in Deepsouth, in that it presupposes that all constituent parts of the invention can and indeed must be localized to the country of protection. The argument seems to originate from what is more commonly known as the “all elements rule”, which makes it clear that for there to be direct patent infringement, an infringing product or service must include each and every element of the patent claim.

In contrast, courts in other countries such as Germany and the United Kingdom have considered more expansive readings of patent statutes and patent claims as a legal necessity to avoid legal loopholes for patent enforcement in cross-border contexts. In the German case Prepaid-Karten II, 2 U 51/08, the Court of Appeals of Düsseldorf found that there was direct infringement in Germany of a method claim, even if all steps but one were performed outside Germany. Instead of dissecting where each claim element was performed, the Court focused on whether there was nevertheless commercial intent to exercise the invention in Germany. In the United Kingdom, courts have instead adopted a claim-focused approach, where what counts as infringing use of an invention within the United Kingdom depends on a construction of what constitutes the “substance” of the invention. See Menashe Business Mercantile Ltd v. William Hill Org. Ltd, [2002] EWCA Civ 1702, and Research in Motion UK Ltd v. Motorola Inc, [2010] EWHC 118 (Pat). This is not too different from how the United States Court of Claims considered several decades earlier in Decca, Ltd. v. United States, 544 F.2d 1070 (Ct. Cl. 1976), that there was infringement in the United States where the invention was “controlled” by the defendant within the United States, where also the “patentees’ true contribution” was deemed to be located.

In sum, while there is a growing trend in several jurisdictions towards engaging in what are more far-reaching practices of localizing infringements, these forms of extraterritorial application have been inconsistent. These tests have also remained largely unclear and unsubstantiated. For some inventions, such as problem inventions and combination inventions, it is not possible to attribute a particular essential part of the invention to a particular claim element. There is no single claim element to construe as the “substance” of the invention. Moreover, with patent infringement being a strict liability tort, there is no plausible legal basis in patent law for introducing a requirement that there must be commercial intent for there to be infringement, let alone in cross-border situations.

Can we find any unifying legal principles that can be extended across jurisdictions? At the end of the day, whether a particular conduct falls under a particular patent claim or not is a technical test. This fundamental criteria for assessing patent infringement should not be abandoned merely because the conduct is now cross-border. Just as we examine when there is infringement from a technical standpoint, so should we evaluate where an infringement occurs from that same lens. This is possible to do without dissecting claim elements, by instead focusing on where the technical contribution of the invention as a whole is achieved. This analysis is however still largely one-sided from a contacts point of view, as it is limited to contacts linked to the relevant country of protection, which is typically the forum. After defining the contacts to the relevant country of protection, it should also be deemed necessary to weigh those contacts established in other countries. Perhaps, and with some effort when adopting a more uniform and balanced legal test, the patent loophole may indeed be closed across borders.

The full article is available to read on the UIC Review of Intellectual Property Law website, here.