CASRIP Newsletter - Spring/Summer 2006, Volume 13, Issue 2
Accommodating New Technologies within Existing Law: Application of the Extraterritorial Infringement Provision 35 U.S.C. § 271(f) to Software Patents
By Vladimir Raskin [*]
American patent law permits imposition of liability for international sales where the product was assembled abroad from components of the patented invention exported from the United States. Liability under this "extraterritorial" provision, codified in 35 U.S.C. §271(f), is imposed for supplying components of a patented invention for assembly overseas, with knowledge that such assembly would be an infringement if done in the U.S. and can be imposed even though the components themselves do not enjoy any patent protection. This provision was enacted in 1984 by Congress as The Patent Law Amendments Act. As software has become an increasingly important player in the U.S. economic and scientific development, application of 35 U.S.C. §271(f) has gained significance and raised many challenges for courts interpreting and applying this provision.
Due to its intangible nature and rapidly growing variety of its use, the unique characteristics of software also make it difficult to fit within some traditional concepts of patent law, such as patentable subject matter and applicable components of an invention. The application of extraterritorial liability for software infringement is gaining importance and raising challenges because, as some legal scholars point out, the rapid increase in global product markets, growth in both domestic and overseas software development, and the increasing trend towards outsourced offshore manufacturing make extraterritorial infringement much more likely.
Furthermore, the increasing variety of software transmission means through the Internet as a channel for software distribution add to the complexity of application of the extraterritorial provision in software patent infringement cases. This article explores the evolution of application of liability under 35 U.S.C. §271(f) to software and highlights some unresolved issues that courts will wrestle with in the future.
II. History of Section 271(f)
Section 271(f) was enacted in response to the famous Supreme Court decision of the case Deepsouth Packing Co. v. Laitram Corp. Deepsouth manufactured the parts of a shrimp deveining machine patented by Laitram in the U.S. and shipped those parts for assembly and use abroad. Although the assembled machines infringed the plaintiff's patents, the Supreme Court held that "it is not an infringement to make or use the patented invention outside of the United States." In other words, it was not an infringement to make all of the components of a patented invention within the U.S. and subsequently ship them abroad for assembly and use. This decision created a loophole that potentially could lead (and some cases, did lead) to serious damage to the patent enforcement and protection system in the U.S.
Twelve years after Deepsouth, Congress acted to close this loophole by amending the Patent Act to add 35 U.S.C. §271(f). Section 271(f) states:
(1) Whoever without authority supplies or causes to be supplied in or from the Unites States all or a substantial portion of the components of a patented invention, where such components are uncombined in whole or in part, in such manner as to actively induce the combination of such components outside of the United States in a manner that would infringe the patent if such combination occurred within the United States, shall be liable as an infringer.
(2) Whoever without authority supplies or causes to be supplied in or from the United States any component of a patented invention that is especially made or especially adapted for use in the invention and not a staple article or commodity of commerce suitable for substantial noninfringing use, where such component is uncombined in whole or in part, knowing that such component is so made or adapted and intending that such component will be combined outside of the United States in a manner that would infringe the patent if such combination occurred within the United States, shall be liable as an infringer.
Effectively, §271(f)(1) imposes liability for those who supply components in a manner that induces infringing combination outside the U.S., and §271(f)(2) imposes liability on those who supply a component that is made or adapted for use in the invention knowing that the combination would result in infringement had it been done in the U.S. Historically, the U.S. patent protection system applied exclusively within U.S. territorial limits. The enacted provision effectively expanded the patent protection outside of the U.S.
Early applications of 271(f) addressed mechanical devices, similar to the Deepsouth case. Later cases involved application of various technologies, including chemical compounds, and attempted to interpret 271(f) more broadly, particularly when the issue concerned what kind of matter could be considered a component for the purposes of §271(f). In W.R. Grace & Co. v. Intercat, Inc., Delaware federal district court expanded protection beyond mechanical components to chemical compounds, thus opening the door for broader interpretation of the "component" element of the statute.
In Standard Havens Products, Inc. v. Gencor Indus., Inc., the Federal Circuit found that §271(f) did not apply to the exportation of an apparatus for performing a patented process, or, in other words, that 271(f) does not apply to method claims. Synaptic Pharm. Corp. v. MDS Panlabs, Inc.,  followed Standard Havens to hold that, while § 271(f) protects against the export of components of patented products, it does not protect against the foreign use of a patented process, even if the unpatented apparatus for performing that process was made and supplied from within the U.S.. The court reasoned that method claims do not have physical "components" capable of being "combined," as required by the plain language of §271(f), into a patented invention. The court stated that the express language singling out process patents for protection in §271(g) indicated that
"Congress knew how to protect against foreign use of process patents, and chose to limit such protection to uses which result in the introduction of products into the U.S...The language and legislative history of §271(f) demonstrate an exclusive focus on the sale of components patented in the U.S. for combination into a finished product, apparatus, or invention abroad."
Thus, based on the express process patent language in §271(g) and lack of language in §271(f) indicating applicability to process patents, courts have declined to apply §271(f) to method claims or design patents based on the absence of "components" that may be assembled abroad to form a patented combination.
Waymark Corp. v. Porta Systems Corp., decided in 2001, shed some additional light on the Federal Circuit's interpretation of §271(f) in regard to the "supply" requirement. The Waymark decision indicated that an act of supplying components abroad, which is not necessarily accompanied by the assembly of a final product, could suffice to impose liability for infringement. The court's reasoning suggested that as an "assembly" was not mentioned in the plain language of §271(f), therefore, Congress did not express the intent to read "assembly" into the statute as a condition of infringement.
Inevitably, similar issues arose apropos of software when the number of software patents filed increased and more software-patent related litigation arose in the late 1990s.
III. Software as Patentable Subject Matter
Patentable subject matter is codified in 35 U.S.C. 101 as any "process, machine, manufacture, or composition of matter, or any new and useful improvement thereof."
The often-quoted Supreme Court decision states that patentable subject matter includes "anything under the sun that is made by man." In 1981, in its Diamond v. Diehr decision, the U.S. Supreme Court opened the door for patenting software by validating the patentable use of software in connection with machinery. The Federal Circuit broadened this protection to include any form of software. In its watershed decision of In re Alappat, the Federal Circuit held that "[c]omputer operating pursuant to software may represent patentable subject matter, as long as claimed subject matters meet all other statutory patentability claims."
Following Alappat, in In re Beauregard, the court authorized protection of software as articles of manufacture, such as software stored on a computer readable medium. And in State Street Bank and Trust Co. v. Signature Financial Group, the Federal Circuit allowed claims for an invention that processes and outputs numbers. State Street Bank rejected the earlier distinction between software and hardware, stating that the relevant patentability test was that the device is patentable as long as it produces "a useful, concrete . . . result."
Thus, software patent protection has gained increasingly broad recognition in the last decade. The law effectively recognizes software as patentable subject matter as long as this software does not represent abstract ideas, formulas or laws of nature. As such, software could be a component of an invention and therefore §271(f) applies to it.
IV. Analysis of Pertinent Case Law
Thus far, U.S. federal district courts have decided only a few cases involving application of §271(f) to software. Interestingly, the majority of cases decided in that area involved Microsoft as a defendant. Even in relatively few decisions, mostly by Federal Circuit which handles the majority of patent-related litigation, the court has managed to significantly develop and, one might say, fundamentally change, the understanding of the scope and application of §271(f) to software.
A. The Enpat Decision: § 271(f) Interpreted Narrowly
In Enpat, Inc., v. Microsoft Corp., Microsoft was accused of infringing a process software patent. The plaintiffs alleged that two Microsoft products, Microsoft Project and Microsoft Team Manager, contributorily infringed, or actively induced infringement of, plaintiffs' patent. The plaintiffs sought to force Microsoft to pay a royalty for its domestic and foreign sales of these products. Microsoft argued that §271(f) applied only to the assembly of patented products abroad and did not apply to method patents, which have no components. This was probably the only time when the court agreed with Microsoft in a case involving application of §271(f). The court stated that if Congress had intended to prohibit U.S. companies from exporting products, which allow foreign companies to make unauthorized use of patented methods, it could have done so in clear, unambiguous language like that found in § 271(g). The court agreed with Microsoft that the language and legislative history of §271(f) demonstrate an exclusive focus on the sale of components patented in the U.S. for combination into a finished product, apparatus, or invention abroad. The court concluded: "While it is true that any process involves the use of physical objects, this alone is not enough to bring a method patent within the purview of § 271(f)." Thus, the court found no §271(f) liability for a method patent with no components where the patent only described steps required to accomplish a task.
B. Pellegrini: § 271(f) Does Not Apply To Instructions for Manufacture
Pellegrini v. Analog Devices, Inc. presented the question whether components that are manufactured outside the U.S. and never physically shipped to or from the U.S. may nonetheless be "supplie[d] or cause[d] to be supplied in or from the U.S." within the meaning of 35 U.S.C. §271(f)(1) if those components are designed within the U.S. and the instructions for their manufacture and disposition are transmitted from within the U.S..
Pellegrini claimed that a line of the defendant's integrated circuit ADMC chips infringed on its patent for a brushless motor drive circuit. Pellegrini contended that "Analog is incorporated in the U.S. and has executive, marketing, and product line responsibilities for ADMC products; that Analog conceived and designed the ADMC products; that Analog is the exclusive manufacturer of ADMC products; that Analog makes all development and production decisions for ADMC products and that "Analog receives purchase orders from and invoices customers worldwide for ADMC products and increases production levels for ADMC products in response to those purchase orders." However, none of the infringing chips were physically manufactured in the U.S..
The court determined that "'suppl[ying] or caus[ing] to be supplied' in §271(f)(1) clearly refers to physical supply of components, not simply to the supply of instructions or corporate oversight." Thus, the court held that, even though Analog may have been "giving instructions from the U.S. that cause the components of the patented invention to be supplied," the fact that the components were not physically present in the U.S. prevented liability from attaching under §271(f)(1). Accordingly, the Pellegrini decision clearly emphasized the physical transmission of components from the U.S. overseas as a prerequisite to attaching liability under §271(f).
C. Imagexpo: The First § 271(f) Case Involving Microsoft
In Imagexpo, L.L.C. v. Microsoft Corp., Imagexpo contended that Microsoft violated §271(f) by exporting NetMeeting software code. Microsoft developed the NetMeeting computer code within the U.S. The code was then exported overseas on "golden master" disks "for the express purpose of combining the code with other components (namely CD-ROMs and computers) to form apparatuses that infringe the patent-in-suit." The NetMeeting software code that is incorporated into the infringing apparatus constituted an exact copy of what was on the "golden master."
Microsoft argued that each template is nothing more than a plan, design, or concept. The template itself is not a tangible item that physically becomes a constituent part of a protected apparatus. Microsoft's position was that the "golden master," or template, is analogous to the shoe sole design patent that was found to be outside the scope of §271 in Aerogroup Int'l. v. Marlboro Footworks, Ltd., 955 F.Supp. 220 (S.D.N.Y.1997).
Relying on the decision in Enpat, Microsoft maintained that "the term 'component' denotes a tangible, physical element of a patented device, commonly associated with an apparatus claim." In Microsoft's view, since the "exported software code is a template, it merely exports the template itself and not some tangible, physical object."
Imagexpo's position was that the "golden master," or template, at issue in this case actually involved an information- or code-base which becomes an integral ingredient in the finished computer product. In other words, the overseas replicator does not simply construct the computer product using a plan, design, or recipe supplied by Microsoft. Instead, the functional nucleus of the finished computer product is driven by the code, which is transmitted through the "golden master." The template is, therefore, a component.
The court agreed with Imagexpo's arguments and denied Microsoft's motion to bar damages under §271(f). The court decided that "the 'golden master' and the electronic codes supplied by Microsoft to its overseas representatives constitute[d] 'components' under 35 U.S.C. §271(f)."
D. Eolas: "The Great Leap Forward"
In Eolas Technologies, Inc. v. Microsoft Corp., decided (on appeal) in 2005, Eolas sued Microsoft for patent infringement, alleging that Microsoft had incorporated Eolas's invention into Microsoft's Internet Explorer. Microsoft exported a limited number of golden master disks containing the software code for the Windows operating system including Internet Explorer to Original Equipment Manufacturers (OEMs) abroad who used that disk to replicate the code onto computer hard drives for sale outside of the U.S. Microsoft argued that the golden master, which itself was not installed on any computer system, was not a component of the patented invention within the meaning of §271(f). The district court determined that the code on the golden master disks constitutes "components" of an infringing product for combination outside of the U.S. under §271(f).
To determine whether software is a component for §271(f) purposes, the court analyzed the language of the statute and concluded that §271(f) was not limited to "patented 'machines' or patented 'physical structures"' or to "'machine' components or 'structural or physical' components." Thus, the software code claimed in conjunction with a physical structure, such as a disk, fits within at least those two categories of subject matter within the broad statutory label of "patented invention." The court further stated that "every form of invention eligible for patenting falls within the protection of section 271(f)" and "every component of every form of invention deserves the protection of section 271(f)."
After concluding that § 271(f) applied to "every form of invention" and "every component of every form of invention," the court determined that software qualified as a patentable invention, both as a process type and product type invention. Next, in deciding whether software code on a golden master is a "component," the court stated:
"A "component" of a process invention would encompass method steps or acts. A "component" of an article of manufacture invention would encompass a part of that construct. Because a computer program product is a patented invention within the meaning of Title 35, then the "computer readable program code" claimed in the patent is a part or component of that patented invention."
Moreover, the court noted that it "accords the same treatment to all forms of invention" and that there is no "principled reason for treating process inventions different than structural products." The court noted that software converts its functioning code into hardware and vice versa. So, in the context of this patented invention, the computer transforms the code on the golden disk into a machine component in operation. Thus, sound policy counsels against varying the definition of "component of a patented invention" according to the particular form of the part under consideration, particularly when those parts change form during operation of the invention as occurs with software code.
Finally, the court clarified its decision in Pellegrini as applied to software. The Eolas court stated that "the language of section 271(f) does not impose a requirement of 'tangibility' on any component of a patented invention." The court further clarified its holding in Pellegrini, stating that "Pelligrini requires only that components are physically supplied from the U.S.. Pelligrini does not impose on section 271(f) a tangibility requirement that does not appear anywhere in the language of that section." In effect, the Federal Circuit refused to make a distinction between intangible software and software recorded on a computer readable medium and suggested that any aspect of the patented invention -- such as a step in a process - could be a component of a patentable invention.
The Eolas decision seemed to have revolutionized the interpretation of 271(f). First, it reaffirmed the Imagexpo holding that software on a master disk should be considered as a component of an invention for the purposes of 271(f) and provided a sound analytical basis for this holding.
Second, by including process patents under protection of 271(f), the Federal Circuit opinion in Eolas appears to have overruled the federal district court decisions in Enpat and Synaptic Pharmaceutical. Although the court did not cite or even mention those decisions, it nevertheless explicitly stated that §271(f) applies to process patents. The court also specifically stated that software claimed as a process is patentable, and indicated that it would be covered by §271(f). Therefore, §271(f) infringement liability would attach for exporting software covered by a process patent.
Third, the Eolas court, by stating that "every form of invention eligible for patenting falls within the protection of §271(f)" may have included design patents under §271(f) protection. It remains to be seen whether the future court decisions will interpret Eolas this broadly.
Fourth, the court, while clarifying Pellegrini, upheld the Pellegrini "bright line rule" that components must be "physically supplied from the United States." This requirement caused electronic transmissions to fall outside the scope of §271(f), giving rise to a "mode of transmission" loophole whereby copiers could export their software and virtual components without fear of infringement liability. The Eolas court did not address the issue of whether exporting software by using electronic transmissions, perhaps over the Internet, would give rise to §271(f) liability.
E. AT&T: Expansion of Eolas
The AT&T Corp. v. Microsoft Corp., decision of 2005 expanded the Eolas interpretation of applicability of §271(f) to software. In this case, AT&T sued Microsoft for allegedly infringing an AT&T patent by selling products containing speech signal compression and decompression software. Microsoft supplied a number of master versions of the Windows software to foreign computer manufacturers and authorized foreign "replicators," to replicate the master versions in generating multiple copies of Windows for installation on foreign-assembled computers that were sold to foreign customers. The master versions were created in the U.S. and were sent abroad on so-called golden master disks or via electronic transmissions.
The master versions of exported Windows incorporated certain speech "codecs," which allegedly infringed AT&T's patent. On appeal, Microsoft argued that the district court erred in its determination of infringement under §271(f), insisting that the master versions of the Windows software that it exports for copying abroad are not "components" within the meaning of §271(f). It also argued that liability under §271(f) should not attach to the copies of Windows made abroad because those copies were not "supplied" from the U.S..
The court quickly disposed of the first issue stating that the question of whether software may be a "component" of a patented invention under §271(f), was answered in the affirmative in Eolas Techs. Inc. v. Microsoft Corp. (Eolas was decided while Microsoft's appeal in AT&T was pending). The court reiterated its statement made in Eolas that "[w]ithout question, software code alone qualifies as an invention eligible for patenting," and that the "statutory language did not limit §271(f) to patented 'machines' or patented 'physical structures,' " such that software could very well be a "component" of a patented invention for the purposes of §271(f).
The remaining question, whether software replicated abroad from a master version exported from the U.S. with the intent that it be replicated may be deemed "supplied" from the U.S. for the purposes of §271(f), was resolved by the court based on the statutory interpretation of §271(f) as applied to software. To properly construe the "supplie[d] or cause[d] to be supplied in or from the United.States" requirement, the court looked at the way software is typically "supplied." Given the nature of the technology, the "supplying" of software commonly involves generating a copy. Copying, the court stated, is part and parcel of software distribution. Accordingly, for software "components," the act of copying is subsumed in the act of "supplying," such that sending a single copy abroad with the intent that it be replicated invokes §271(f) liability for those foreign-made copies. Thus, Microsoft was held liable for each act of copying of its master disk done abroad. The court stated that the mode of supplying, i.e. whether software is sent abroad via electronic transmission or shipped abroad on a master disk, is a distinction without a difference for the purposes of § 271(f) liability. "Liability under § 271(f) is not premised on the mode of exportation, but rather the fact of exportation."
Significantly, the court fortified its decision by referencing the public policy behind §271(f): "It would be unsound to construe a statutory provision that was originally enacted to encourage advances in technology by closing a loophole, in a manner that allows the very advances in technology thus encouraged to subvert that intent. Section 271(f), if it is to remain effective, must therefore be interpreted in a manner that is appropriate to the nature of the technology at issue."
Thus, the court clarified and expanded its Eolas decision by stating for the first time that the mode of supplying does not matter for the purposes of § 271(f), effectively closing the Eolas "mode of transmission" loophole. The main holding of the case, that copies made entirely overseas from a single version supplied from the U.S. were separate acts of infringement, was a significant development in §271(f) interpretation as it opened the door for potentially multiple recoveries for the same act of infringement. The AT&T court appears to have implicitly overruled the Standard Havens decision that §271(f) did not apply in the case of the exportation of an apparatus for performing a patented process. The key to the court's decision was its conviction that the statute must evolve to take into account new technologies, such as the fact that software is easy to copy.
F. Union Carbide, Last Case of "271(f) Trilogy": Further Expansion of Eolas
In Union Carbide Chems. & Plastics Tech. Corp. v. Shell Oil Co., also decided in 2005, the court further expanded the Eolas and AT&T analysis, although the case itself concerned chemical compounds rather than software. The court held that the catalyst called for in a patent for the process of making ethylene oxide, sent abroad by an infringer for use by an affiliated foreign chemical producers, was a "component" of a patented invention, within meaning of the statute proscribing foreign assembly.
As the court stated, the phrase "any component of a patented invention," in the statute proscribing foreign assembly of patent-infringing goods using U.S.-supplied components, applies to components used in the performance of patented process and method inventions. In other words, manufacturing products overseas via a process, patented by the other, which uses a product (component) supplied from the U.S., constitutes infringement of the U.S. patent. This decision manifested the further acknowledgement that the process patents can be the subject of §271(f) inquiry.
V. Where Are We Now?
2005 was a significant year for developments in the interpretation of §271(f) and its application to software patent infringement cases. Courts have adapted an increasingly broad view of how §271(f) applies to software, despite the cautionary warnings expressed by some legal scholars.
Courts have clarified the notion of a "component" as it applies to software under §271(f). First, a component can be anything that is, in some broad sense, "part" of an invention. Second, a component need not be something "physical" or "tangible." Third, a component need not be a claim element. Fourth, the component supplied for assembly overseas from the U.S. territory need not be a part of the product manufactured overseas. For the purposes of §271(f), it is sufficient if the component is somehow used (in a broad sense, for example, used in performance of inventions) to make the overseas product. Moreover, the "idea" for an invention itself can be considered as a "part" of the invention and therefore will fall within the expanded definition of component.
Courts indicated that the actual mode of transmission or "supply" is irrelevant: for purposes of §271(f), the copying of a "component" overseas is equal to the supply of the component from the U.S. Finally, courts have expanded the scope §271(f): it appears that §271(f) liability will now attach to process patents and, arguably, to design patents.
VI. Lingering Questions
Despite the spectacular progress made in the area of liability for extraterritorial infringement of software patents many issues still remain.
A. Is Pellegrini Reconcilable with Eolas-AT&T-Union Carbide?
First, despite the Federal Circuit's clarification of Pellegrini in its Eolas decision, it may still be difficult to reconcile Pellegrini and the subsequent Eolas-AT&T-Union Carbide trilogy. In Pellegrini, the patent claim was not directed to software but to a computer chip. However, software was used to an extent that instructions in a software form were sent overseas and used to manufacture a patented computer chip. In Eolas and AT&T the software itself was part of a patent. It seems that the difference between the patented chip manufacture instructions in a software form and a software that is a subject matter of a patent is fairly subtle. However, Pellegrini and the subsequent trilogy of cases came out differently.
In Union Carbide, the product manufactured overseas via a process that used a catalyst supplied from the U.S. constituted infringement of the U.S. patent. In both Pellegrini and Union Carbide something was sent overseas to make computer chips (Pellegrini) or product (Union Carbide); but is there any difference between "instructions" used to make the infringing product (in Pellegrini, no infringement found) and the "component" used somehow in manufacturing the infringing product (in Union Carbide, infringement was found)? According to Union Carbide, the accused component need not be part of the accused product. Could then the component fall under the "instructions" category?
B. The "Physical Supply" Requirement: Should It Be Left Intact?
The second issue is the Federal Circuit's reading of Pellegrini, which indicates that, while components themselves do not have to be physical, they do have to be "physically supplied" from the U.S. (no matter what mode of supply took place, as was clarified in AT&T). Nowhere in the AT&T or the Union Carbide did the court explicitly overrule this reading of Pellegrini. Leaving the "physical supply" requirement intact poses some questions. For example, what exactly was "physically" supplied in the AT&T case? Was it the code on a master disk that was copied numerous times abroad? Did the actual copying process contain any elements physical elements or components supplied from the U.S.?
The copying involved the use of electromagnetic signals simulating "zeros" and "ones" of the copied machine code. These signals were not produced in the U.S., but rather were generated in the machines that were used for the copying process overseas. Technically speaking, nothing was actually "physically supplied" from the U.S. territory. Admittedly, this is a fairly farfetched argument, but it yet could be a legitimate one provided that the courts change their view on whether the machine code in the form of an electromagnetic signal is patentable subject matter.
C. Unlimited Liability?
The third issue is a somewhat uneasy conclusion that one may draw from reading the §271(f) trilogy of cases. If a specific manner of supply (by electronic transmission or otherwise) makes no difference, then any software exported from the U.S. could potentially lead to a patent infringement under §271(f). Even a source code paper printout brought from the U.S. overseas would potentially create liability for infringement. Such a protection regime might potentially lead to unlimited liability as any software idea developed in the U.S. and somehow used in a patent-infringing product overseas could lead to worldwide infringement. In practice this may mean that liability may not be limited to just the U.S. recovery for the plaintiff; it could also lead to a possibility of recovery in each individual country where the infringing product is being produced. With software products easily distributed throughout the world, such a liability regime could create a disincentive for companies in emerging technology areas to base their development operations in the U.S.
One could speculate more about potential consequences of the §271(f) liability regime adopted by the court. For example, in the pharmaceutical area, the new regime could lead to a massive renewal or revision of existing licensing obligations, because sending a chemical formula overseas for use in a potentially infringing product can amount to §271(f) infringement.
However, none of those concerns have so far materialized.
Most importantly, patent law must be able to adapt to rapid changes in technology. The way products and services are designed, manufactured and distributed is vastly different from what it used to be: almost all involve computers and computer-related products, including software. Congress and the courts are trying to keep up with these profound changes and the application of liability for extraterritorial infringement to software and its components is just one example of this difficult, challenging work. Keeping in mind the "big picture," courts' application of §271(f) to software infringement cases is the work that needs to be done and is being done to make sure inventors' patent rights are protected and are not diminished.
Section 271(f) was supposed to and should be interpreted to provide protection against the exportation of software and its components. As long as the application of §271(f) to software infringement cases does not cause an opposite result, it should serve to protect American inventors from extraterritorial infringement in software field.
Indeed, the recent court decisions are more likely than not to be a step in the right direction for American inventors. The "doomsday" scenarios proposed by Microsoft are yet to materialize.
Nonetheless, the last word in the §271(f) "saga" has yet to be spoken. There are proposals to repeal §271(f) altogether as part of an upcoming patent reform in the U.S. In addition, there likely be more cases on the subject that have yet to be decided. There is also a possibility that the U.S. Supreme Court will grant certiorari to the AT&T case and will render its own vision of how §271(f) should be applied to software infringement cases. Thus, the legal and inventor communities are expecting further developments in the complex and challenging area of prevention of extraterritorial infringement of software patents.