Intellectual Property Bulletin Fall 2006

Potential Defenses of Implied Patent License under the GPL


The GNU General Public License v2.0 ("the GPL") governs a majority of open source software. The GPL grants recipients the right to copy, distribute and modify covered software provided they do so in a manner consistent with the "free" nature of the software. Anyone can be an eventual recipient of software released under the GPL and the distributor cannot restrict who will eventually receive copies. The GPL requires any distribution of a software program and derivative works based on the program to be licensed under the GPL. Thus, the GPL addresses copyright issues but does not contain a clear patent license grant.

The previous issue of this Bulletin (Summer 2006) included an introductory article, which discussed whether open source and patent rights can coexist. In general, patent rights may be substantially limited due to an implied license when the target infringing activity is covered by the GPL. On the other hand, infringing activity that falls outside the GPL scope may be subjected to a patent infringement suit.

This follow-up article discusses in more depth the impact of the GPL on the patent rights of the patentee, and various implied license theories that users of GPL'd software may be able to employ in defense against a patent suit. In discussing each of the implied license theories, reference is made to the hypothetical case where a company releases software under the GPL and then sues a recipient of the software for infringement of a patent that covers the software.

Implied License Theories

According to Federal Circuit precedent, there are four doctrines of implied license, each of which has a different set of requirements: legal estoppel, equitable estoppel, conduct and acquiescence. Typically one of the two estoppel theories is used to imply a license; the more abstract conduct and acquiescence theories are used rarely but still acknowledged as valid. See Wang Lab. v. Mitsubishi Elecs. Am., 103 F.3d 1571 (Fed. Cir. 1997). This theory prevents a patentee/licensor from acting to take away from a right that the licensor already agreed to grant the licensee. In determining the existence of legal estoppel, a court considers the scope of the right granted and whether the patentee's actions impact those rights. Where the remedies sought by the patentee would not impact what is granted under the license, there is no reason to imply a license through legal estoppel.

In the hypothetical case above, the accused infringer has an argument that the elements of legal estoppel are met. Through the GPL, the patentee has granted the accused infringer the right to make and distribute the invention. While usage rights may not be explicitly granted under the GPL, it can be argued that they are necessary to effectively practice the rights that have been granted. Both the right to practice the invention (through the GPL) and an attempt to derogate that right (by claiming the licensee has no right to use the licensed invention) are present.

Satisfying the legal estoppel theory also requires showing that the licensor received valuable consideration for the license. One possible item of consideration received under the GPL is the reciprocity agreement — the promise by the licensee to license any further distribution of the program and any works based on it under the terms of the GPL. In Wang, the proliferation of the plaintiff's technology and adoption of it as an industry standard were enough to form consideration under legal estoppel. The licensee may be able to argue that the benefits any licensor receives from agreeing to comply with the licensee form sufficient valuable consideration to imply a license by legal estoppel.

Equitable estoppel uses a very different theory to arrive at an implied patent license. This defense involves finding a reasonable inference by the licensee based on the licensor's conduct. Equitable estoppel requires that the patentee communicate to the accused infringer in some manner its intent not to sue, that the infringer rely on this communication, and that allowing the patentee to proceed with its claim would materially prejudice the alleged infringer. As with the other implied license doctrines, fairness and factual details factor heavily into the finding.

The accused infringer in the hypothetical case above could argue that the patentee's distribution of the software under the GPL was the conduct that communicated intent not to sue; it would be reasonable for a user to infer that a product released under the GPL can be used without further permission from the author. In choosing to download and run the program, the accused infringer relied on the fact that it was released under the GPL. The licensee should also have little difficulty showing material prejudice. While the typical situation invoking this doctrine involves a much more direct relationship between the parties, the facts of this hypothetical case would most likely allow a court to imply a license under the asserted patent.

Even if the court accepts neither of the above estoppel arguments as a basis for an implied patent license, the patentee's decision to release software under the GPL, allowing people to download and use that software without protest, might support an implied license based on conduct or acquiescence. It is well within the legal framework of the implied license doctrine to make such a judgment if equity so requires.

Implied License for Modified GPL Software

The GPL permits licensees to do more than just copy and distribute the program; it also allows them to modify it and to then release those modifications under the GPL. If the modified versions violate the patent noted in the hypothetical case above, the same implied license arguments may protect the modifiers and any downstream users. A modified version of the program that embodies the same invention as the original version clearly falls within the rights licensed by the GPL.

It is also possible that a later modified version of the program might actually embody the invention of a different patent held by the distributor. This forms a more difficult scenario, because the GPL v2.0 puts no inherent limits on what modifications can be made to the licensed program. Later derivative works may function very differently from the original program that the company released and licensed. A modification may embody patents that the company never intended to allow to be used freely.

Where a company sues on a patent embodied in the changes of a derivative work, the arguments to imply a license may be less persuasive. In particular, each theory of implied license requires conduct by the licensor that indicates an intention not to enforce the patent. Such requisite conduct does not seem present for a patent covering software not included in the patentee's GPL release. Both of the estoppel doctrines have additional requirements that make their application to this example considerably weaker.

In more detail, legal estoppel requires that the patentee attempt to derogate from a granted right. Arguably, however, the right that the patentee granted to modify the program did not extend so far as to cover other patents, and courts have declined to extend analogous doctrines (such as the first sale doctrine) to modifications that infringe separate patents not covered by the unaltered invention. Despite the right granted under the GPL, it is possible that legal estoppel will not protect this additional conduct. Note that some drafts of the GPL v3.0, which has not been finalized when this article was written, expressly state that the licensor grants a patent right in patents covered by certain types of future-made derivative works.

Similarly, equitable estoppel requires reliance on communication made by the patentee to the accused infringer. Thus, protection of a program modification would require knowledge by the accused infringer (modifier) that the modifications infringed the patent of an upstream patentee, and belief by the modifier that the GPL reflected the patentee's intent not to enforce that patent against a modified version of the program. It is assumed that such factual findings would rarely occur or be provable.

In short, the conduct of releasing software under the GPL, while implying permission from the patentee to use this specific software and alter or improve the software in customary ways, does not necessarily imply permission to add to the software other separate inventions patented by the patentee. Specifically, given the language of the current GPL v2.0, it is certain that a modified version of software that infringes patents covering other inventions not released under GPL by the patentee will not provide its users with the same implied license defenses likely accessible to those who only practice inventions available in the original software released under GPL by the patentee.

Broad Judicial Discretion

The existence of an implied patent license is a question of law, and an implied license is an equitable defense to patent infringement. A court considers the totality of the relevant circumstances in determining whether an implied license is warranted, and the specific facts-in-question and fairness considerations are always relevant to the court's decision.

Because the GPL is different from licenses normally available to business entities under patent law, and the applicable equitable doctrines are under broad judicial discretion, the courts may ultimately dispense with GPL issues in a manner different from that discussed herein.

But from available case law, it is reasonable to conclude that the implied license defense is available and tenable for a defendant in a patent suit involving software released under the GPL.

The Patent Prosecution Highway: A First Step for International Patent Harmonization?


Over twenty years ago, the Trilateral Co-operation was set up between the United States Patent Office (USPTO), the European Patent Office (EPO) and the Japan Patent Office (JPO), with a lofty goal: "to contribute to an increasingly efficient worldwide patent system in the 21st century." (See One of the long-term goals of this cooperation is to provide full mutual exploitation of search results among the Trilateral Offices. In July of this year, the Trilateral Offices took a first step toward this goal, with the introduction of a pilot program called the Patent Prosecution Highway.

The Patent Prosecution Highway (PPH) is a cooperative effort between the USPTO and the JPO that leverages accelerated examination procedures available in both offices and allows each office to benefit from examination previously done by the other. In short, the PPH allows an applicant/company in the office of first filing to use a simplified procedure to request accelerated examination in the office of second filing. (See In conjunction with this process, prosecution documents from the office of first filing (including office actions, references cited and patentable claims) are substituted for the pre-examination search and other documentation required for traditional accelerated examination. Currently the PPH is slated for a one-year trial period; however, it may be extended for an additional year. In order to evaluate the program, the Trilateral Offices ask that each applicant/company participating in the PPH program during the trial period submit an evaluation of the program.

A closer look at the requirements for applications eligible to participate in the PPH reveals that the program applies to only a small percentage of patent applications, and that it is more beneficial to applicants who first file in the JPO and subsequently file in the USPTO.

First, the program applies only to a utility application subsequently filed in one of the two offices that claims priority to a first utility application filed in the other office under the Paris Convention. This means that all applications that reach the second office via the Patent Cooperation Treaty (PCT) are ineligible, as are all plant and design applications, reissue applications and re-examinations. For direct national filings, seven times as many patent applications first filed in the JPO later enter the USPTO under the Paris Convention (56,000) each year as do USPTO-filed patent applications that later enter the JPO via this route (8,000 per year). (See "An Intellectual Property-based Nation and the New Route,'" Presentation of JPO Commissioner Makoto Nakajima at the Trilateral Users Meeting, Munich (Nov. 17, 2005), One reason for the disparity in these numbers is that 64 percent of U.S. patent applications that are later filed in the JPO are filed using the PCT process (and thus are ineligible for the PPH), whereas only 8 percent of Japanese patent applications that are later filed in the USPTO use the PCT.

Another requirement that limits the number of applications to which the pilot program will apply is that at least one claim has to have been allowed in the office of first filing, but the application pending in the office of second filing cannot have begun examination. Based on the lengthy backlog at the USPTO in many art units as compared to the backlog at the JPO (average 31 months to final decision), this too makes it less likely that applications first filed in the USPTO will benefit from the program.

Despite these limitations, the PPH may hold some long-term value for U.S. based applications/companies. Instead of focusing on the current PPH pilot program in a vacuum, the program should be considered in the context of the long-term goal of the Trilateral Offices, which is to provide full mutual exploitation of search results. Since the pilot program is slated to run for just one year, the pool of applications eligible for the program may be small. However, the number of applications participating in the PPH may prove useful in consideration of the fact that it is a pilot program — generally considered an evaluation or sampling process used to identify and correct problems before larger-scale implementation. Only time will reveal the efficacy of the program in substance, and as a model for future cooperation.

The study of mutual exploitation of search results as a whole is not new to the Trilateral Offices; the Offices have been evaluating the usability of searches performed by the other offices for over five years. In addition, the Trilateral Examiner Exchange Program has allowed examiners reciprocally to be sent to another patent office since April 2004, and in October 2004, a Dossier Access System went into operation that makes examination documents available online among the Trilateral Offices. Moreover, six conferences have been held since 2000 to deliberate international harmonization of substantive aspects of patent law. Thus, the PPH can be seen as a step toward the goal of mutual exploitation of search results among the Trilateral Offices.

It has been estimated that the Trilateral Offices account for about 85 percent of global patent applications. Among these, more than 200,000 applications are duplicated. Thus, a program of mutual exploitation of patent examination could dramatically decrease redundancy in examination between the Trilateral Offices and could result in decreased backlogs in all three offices. As a result, many believe that the Trilateral Offices could be the ones to spur harmonization of international patent law. Of further note, the Trilateral Offices are not alone in focusing on the issue of international cooperation. In September, 43 countries agreed on a standard system for the patent approval process. (See The agreement has not yet been finalized; however, the working group formed will focus on selected issues such as novelty provisions, the definition of prior art and grace periods for international filings. Seen in this context, the PPH may turn out to provide a foundation for international patent harmonization in the future.

Quick Updates

Federal Circuit Continues Fervent Defense of "Motivation to Combine" Test

Clearly not losing sight of the upcoming Supreme Court oral arguments in KSR v. Teleflex, the Federal Circuit again added to the recent flurry of cases defending its "motivation to combine" test for showing obviousness. In Dystar Textilfarben v. C.H. Patrick (Fed. Cir., Oct. 2006), the court devoted much of its opinion to guard the motivation test and refute "notions put forth recently by various commentators and accepted in major reports."

Dystar concerned a dispute over a patent covering an improved process for dyeing textile materials. The alleged infringer defended that the asserted patent claims were obvious in view of various prior art references. Quickly turning its obviousness analysis to the question of motivation, the court made it clear that motivation to combine references "need not be found in the prior art references themselves, but rather may be found in 'the knowledge of one of ordinary skill in the art, or, in some cases, from the nature of the problem to be solved.'" The court cast attempts to require that motivation be present in the references as being based on "isolated dicta" and resulting from a lack of  "careful, candid, and complete legal analysis." Given the full body of the court's precedents, the proper motivation test, the court held, is not rigid but quite flexible and necessarily involves "consideration of common knowledge and common sense."

As spelled out by the court, the "common sense" approach for finding an implicit motivation to combine may arise out of a desire to have a product or process that is "stronger, cheaper, cleaner, faster, lighter, smaller, more durable, or more efficient." The court went on to say that because "the desire to enhance commercial opportunities by improving a product or process is universal... there exists in these situations a motivation to combine prior art references even absent any hint of suggestion in the references themselves."

To support its construction of the motivation test, the Dystar Court cited various Federal Circuit and Supreme Court cases. In Pro-Mold & Tool Co. v. Great Lakes Plastic, 75 F.3d 1568 (Fed. Cir. 1996), the Federal Circuit found that no documentary evidence was needed to show that it would have been obvious to design a sports trading card holder to be only slightly larger than the trading card because "mankind, in particular, inventors, strive to improve that which already exists." In Mazzari v. Rogan, 323 F.3d 1000 (Fed. Cir. 2003), the Federal Circuit held a patent covering the use of underwater acoustic waves to kill zebra mussels to simply be "more efficient" and thus obvious in view of a reference disclosing the use of an acoustic wave generator and another reference disclosing a method of using water-borne acoustic waves to kill zebra mussels. The Federal Circuit also pointed to the Court's decision in Ag Pro v. Sakraida, 425 U.S. 273 (1976), in which the Court found a patent for a barn designed to send water cascading through the dairy and sweeping waste to downhill drains to be obvious because the purported invention was merely "the work of [a] skillful mechanic" exploiting the principle of gravity.

While firmly defending the position of its well-established motivation test for showing obviousness, the Federal Circuit at the same time has seemingly heightened the threshold to be met for establishing nonobviousness. The court, specifically dismissing reports by the Federal Trade Commission and the National Academy of Sciences viewed by many as advancing current patent reform objectives, sets forth that a motivation to combine exists wherever a sensible reason to combine, explicit or not, is present in the references or based on common knowledge.

Ninth Circuit Reaffirms Vitality of Laches as a Defense to Injunctive Relief in Trademark Infringement Cases

Last month, in the case of Tillamook Country Smoker, Inc. v. Tillamook County Creamery Ass'n (9th Cir., Oct. 2006), the Ninth Circuit affirmed, for the third time in three years, a district court's refusal to enter an injunction against an alleged trademark infringer on the ground of laches, an equitable defense "derived from the maxim that those who sleep on their rights lose them." In doing so, the court highlighted the dangers of failing to aggressively assert trademark rights against a known user of a potentially confusing mark.

Tillamook County involved a dispute between a cheese- and butter-making collective that marketed its products under the "Tillamook" name and a processed meat manufacturer that began selling similarly branded products years later. While acknowledging evidence of actual consumer confusion between the two marks, including the Trademark Office's refusal to register the junior user's mark because it was very similar to the senior user's mark, the Ninth Circuit determined that under the factors it first articulated in E-Systems v. Monitek, 720 F.2d 604 (9th Cir. 1983),  "the meat people" were entitled to continued use of their mark.

As with all laches claims, the court focused in large part on the length of the senior user's delay in bringing suit. In the trademark context, the delay is measured from the time the senior user "knew or should have known" that it had "a provable infringement claim against the [junior user]." In Tillamook County, that delay was 25 years. In Grupo Gigante v. Dallo & Co., 391 F.3d 1088 (9th Cir. 2004), decided two years earlier, plaintiff's delay of four years in bringing its infringement suit was determined to be sufficient to establish laches under the totality of the circumstances. Prejudice to the defendant, the other essential element of a laches claim, is satisfied where a junior trademark user proves "that it has continued to build a valuable business around its trademark during the time that the plaintiff delayed the exercise of its legal rights." Grupo Gigante, 391 F.3d at 1105.

As part of its analysis, Tillamook County reaffirmed the principle of "progressive encroachment," which permits an otherwise dilatory trademark owner to nevertheless prevail in its infringement action if it can demonstrate that the alleged infringer "expanded its business into different regions or into different markets." In doing so, however, the Ninth Circuit emphasized the narrowness of the doctrine, holding that the growth of an existing business is not sufficient to invoke it. Under the facts of the case before it, the Ninth Circuit held that the recent decision of meat manufacturers to begin selling their products in supermarkets did not qualify as "encroachment," but hypothesized that a decision to begin selling cheese products may do so.

DMCA's Anti-Circumvention Provisions Not Enforceable Against Federal Government

The Court of Federal Claims has determined that even though the federal government is subject to copyright infringement lawsuits, sovereign immunity shields the government from a claim that it violated the Digital Millennium Copyright Act's anti-circumvention provisions.

In Blueport Co. v. United States, 71 Fed. Cl. 768 (Fed. Cl. 2006), a software vendor provided the Air Force with a program that included an automated expiration function designed to prevent unlicensed use. When the Air Force allegedly hacked the program to allow use beyond the original license expiration date, Blueport brought suit. Blueport alleged that, among other things, the hacking was in violation of the provision in the DMCA prohibiting any person from "circumvent[ing] a technological measure that effectively controls access to a work protected under [copyright law]." The government moved for summary judgment, asserting that it enjoyed sovereign immunity from such a claim.

The sovereign immunity doctrine has long been applied by U.S. courts to prevent any suit against the federal government, excepting only those for which the government has expressly waived such immunity. In this case, U.S. copyright law expressly waives immunity such that the government can be sued for copyright infringement. Indeed, Blueport's first count against the government was for copyright infringement, and the government did not assert immunity with respect to that count. The government brought its summary judgment motion arguing that even though the DMCA is part of U.S. copyright law, the Copyright Act was not amended to specify that circumvention under the DMCA, in addition to copyright infringement, is within the waiver of sovereign immunity.

From the outset, the court's analysis did not bode well for Blueport. "It has often been said that one of the hardest things in the world is to sue the federal government." Continuing, the court noted that it is "black letter law" that in the absence of its consent, the government is immune from suit, and such consent "cannot be implied but must be unequivocally expressed."

The court held that the anti-circumvention provisions of the DMCA did not expand the definition of what constituted copyright infringement, but gave birth to a new violation — one that was not covered by the existing waiver language in the Copyright Act. The court held that other references within the Act, which Blueport contended showed a congressional intent to waive immunity more generally, were not sufficiently clear.

There has been a great deal of attention given to sovereign immunity issues since a bare majority of the U.S. Supreme Court in 1999 held unconstitutional the laws subjecting states to patent and trademark infringement. Florida Prepaid Postsecondary Educ. Expense Bd. v. College Sav. Bank, 527 U.S. 627 (1999) and College Sav. Bank v. Florida Prepaid Postsecondary Educ. Expense Bd., 527 U.S. 666 (1999). Several bills were introduced after those decisions that would have required some waiver of sovereign immunity in order for a state to be allowed the full benefit of its own intellectual property portfolio. To date, though, no such legislation has been enacted.

The Blueport case is a stark reminder to technology companies that doing business with the government requires a different set of rules and strategies from day-to-day interactions with organizations in the private sector.

Online Journalists Not Compelled to Reveal Trade Secret Leak

In 2004, Internet news sites published trade secret information about an unreleased Apple Computer product called "Asteroid" that facilitates production of digital audio recordings. In response, Apple brought suit for trade secret misappropriation against 25 unnamed "Doe" individuals in California who had allegedly leaked this information to editors/publishers of,, and Mac News Network. Apple Computer v. Does, Case No. 1-04-CV-032178 (Cal. Super. Ct. 2005). To determine the identity of the Does, Apple sought to serve subpoenas on the news sites and their email service providers to obtain emails that might reveal names of the trade secret leak sources. To protect their confidential sources and prevent release of the emails, the news editors/publishers moved for a protective order, which the court denied.

The journalists, represented by the Electronic Frontier Foundation, responded by filing a petition for writ of mandate or prohibition to compel the trial court to set aside its denial of the protective order. In this case, O'Grady v. Super. Ct., modified, 139 Cal.App.4th 1423 (Cal. App., 6th Dist. 2006), the appeals court overturned the lower court ruling that had permitted the Apple subpoenas.

The court considered whether the disclosures sought by Apple were prohibited under the Stored Communications Act (Pub. Law 99-108 (Oct. 21, 1986)) that prevents email service providers from disclosing the contents of a stored message. The court addressed Apple's various arguments, including one that Congress did not intend the Act to preempt civil discovery of stored communications. The statute does not include such an exception, so the court found it should not "lightly engraft exceptions to plain statutory language" but should "let the representative branch of government do its job."

The court further found that petitioners were entitled to protection against disclosure of their confidential sources under California's reporter shield law set forth in the California Constitution. Apple and amici asserted that petitioners should not be protected under the shield law because they were not engaging in "legitimate journalistic purposes," but the court refused to establish any test distinguishing between legitimate and illegitimate journalism. The court took issue with Apple's "pervasive misuse of the verb 'post'" in the briefs filed by Apple and amici characterizing petitioner's conduct as "posting information on a website" as opposed to conducting legitimate journalism. The court concluded that the reporter shield law was not intended to be limited to printed publications, and so petitioners should benefit from its protections.

The court further found that the petitioners should be protected under the First Amendment of the U.S. Constitution as journalists, explaining that the distinctions between online journalists and traditional ink and paper journalists were "minute, subtle, and constitutionally immaterial." The court applied factors established in Mitchell v. Superior Court, 37 Cal.3d 268 (Cal. App., 6th Dist. 1984), agreeing with Apple that the information was not only relevant, but also "potentially crucial" to Apple's trade secrecy claim. On the other hand, the court did not see that all other potential avenues for identifying the sources of the information in question had been exhausted.

Fenwick & West's Intellectual Property Group offers comprehensive, integrated advice regarding all aspects of the protection and exploitation of intellectual property. From providing legal defense in precedent-setting user interface copyright lawsuits and prosecuting software patents to crafting user distribution arrangements on behalf of high-technology companies and implementing penetrating intellectual property audits, our attorneys' technical skills enable the Firm to render sophisticated legal advice.

The contents of this publication are not intended and cannot be considered as legal advice or opinion.

© 2011 Fenwick & West LLP. All Rights Reserved.

We appreciate your feedback!

If you have questions, comments, or suggestions for the editors of the IPB, you can e-mail them to

For subscription requests and address changes, please e-mail