Patent Exhaustion and Implied Licenses: Important Recent Developments in the Wake of Quanta v. LG Electronics

8/4/2009
Client Alert

A little more than a year ago, the U.S. Supreme Court, in Quanta Computer, Inc. v. LG Electronics, Inc.,[1] clarified important questions regarding the application of the patent exhaustion doctrine to system and method patents.  In Quanta, the Court also rejected the notion that patent exhaustion can be avoided by mere restrictive notices to downstream customers where the sale was otherwise authorized by the patent owner.[2]  However, other much debated issues surrounding the law of patent exhaustion were not expressly addressed by the Court, most notably:

  • whether a “covenant • not to sue,” as opposed to a license, amounts to an authorization to sell for purposes of patent exhaustion;
  • whether the “conditional sale” doctrine established by the Federal Circuit in its highly controversial Mallinckrodt decision[3] is still viable; and
  • the questions raised by the Federal Circuit’s Jazz Photo decisions[4] regarding the application of the patent exhaustion doctrine to sales occurring outside the U.S.

The Supreme Court’s silence on these questions left commentators to speculate whether lower courts would still find some guidance in Quanta when confronted with these questions.  In three recent decisions the Federal Circuit, the Eastern District of Kentucky, and the Northern District of California have done exactly that in relying on Quanta in potentially far-reaching decisions addressing each of the controversial issues noted above.

In Transcore[5] the Federal Circuit held that “an unconditional covenant not to sue authorizes sales by the covenantee for purposes of patent exhaustion.”  In addition, and perhaps more remarkably, the court significantly expanded the implied license doctrine, as it is commonly understood based on prior Federal Circuit precedent, by expressly recognizing “an implied license . . . by virtue of legal estoppel” where the assertion of a patent not included in a license grant or covenant not to sue would be “in derogation” of rights expressly granted.

The Federal Circuit’s decision was based on the following factual scenario: In 2000, Transcore and Mark IV Industries, competing providers of toll collection system technology (such as transponders and readers used for E-ZPass) settled a patent infringement suit brought by Transcore.  In exchange for a payment of $4.5 million by Mark IV, Transcore “agree[d] and covenant[ed] not to bring any demand, claim, lawsuit, or action against Mark IV for future infringement” of certain specified patents, including the ‘082, ‘183, and ‘275 patents.

Several years later, ETC, a toll collection system integrator, won a bid to install a tolling system for the Illinois State Toll Highway Authority (ISTHA).  The system itself was purchased by ISTHA from Mark IV.  Transcore sued ETC for infringement of the ‘082, ‘183, and ‘275 patents as well as the ‘946 patent, a patent that had not issued at the time of the Transcore-Mark IV settlement.

The Federal Circuit affirmed the district court’s dismissal of Transcore’s claims based on patent exhaustion, implied license, and legal estoppel.

A Covenant Not to Sue Amounts to Authorization for Purposes of Patent Exhaustion

The Federal Circuit agreed with the district court’s finding that Mark IV’s sale of the toll collection system installed by ETC was authorized by the Transcore-Mark IV settlement agreement and that, as a result, Transcore’s patent rights with respect to the system were exhausted.  Citing Quanta, the Federal Circuit stated the rule that exhaustion is triggered only by a sale authorized by the patent holder, and then answered in the affirmative the crucial question whether “an unconditional covenant not to sue authorizes sales by the covenantee for purposes of patent exhaustion.”

The court cited numerous cases in support of the proposition that a non-exclusive patent license is nothing more than, and equivalent to, a covenant not to sue.  It concluded that the difference between a “covenant not to sue” and a “license” is “only one of form, not substance” and that “both are properly viewed as ‘authorizations’” for purposes of patent exhaustion.  The important question, according to the court, was not whether the settlement agreement between Transcore and Mark IV contained a covenant not to sue or a license, but rather whether it authorized sales by Mark IV.  The court found that sales were authorized, because the covenant not to sue applied to any future infringement and was not limited to, for example, only the acts of making or using.  Consequently, the court found it to authorize all acts that would otherwise be infringing, including the sale of products covered by the specified Transcore patents.

A License May Be Implied by Virtue of Legal Estoppel

Transcore also asserted the ‘946 patent against ETC. This patent had not yet issued at the time of the Transcore-Mark IV settlement.  It was thus not specifically identified in the covenant not to sue. Moreover, the settlement agreement expressly provided that “[t]his Covenant Not To Sue shall not apply to any other patents issued as of the effective date of this Agreement or to be issued in the future.”

Despite the seemingly clear exclusion of patents issuing after the effective date of the settlement agreement, the Federal Circuit agreed with the district court’s finding that Transcore’s rights under the ‘946 patent were exhausted as a result of Mark IV’s sale of the toll system to ISTHA.  With respect to the ‘946 patent, both courts found the sale to be authorized “under an implied license to practice that patent by virtue of legal estoppel.”  Citing its decision in Wang v. Mitsubishi, the Federal Circuit explained that the doctrine of legal estoppel applies “where a patentee has licensed or assigned a right, received consideration, and then sought to derogate from the right granted.”[6] Because the ‘946 patent was broader than, as well as necessary to practice, the ‘082 patent, which was expressly specified in the settlement agreement, the court reasoned that in order for Mark IV to obtain the benefit of the covenant with respect to the‘082 patent it had to be permitted to practice the ‘946 patent to the same extent as the ‘082 patent.  As a result, reasoned the court, Transcore was estopped from asserting the ‘946 patent “in derogation of the authorizations granted to Mark IV under the... ‘082 patent” and Mark IV was an implied licensee of the ‘946 patent. The court disregarded the express language in the settlement agreement that the covenant not to sue “shall not apply to any other patents... to be issued in the future,” stating that this language did not permit Transcore to derogate from the rights it had expressly granted and did not preclude a finding of estoppel.

Finally, because it found Mark IV to have an implied license under the ‘946 patent, the court found its sales to be authorized, and Transcore’s rights exhausted, under that patent.

Significance of the Transcore Decision

The Transcore decision makes it clear that framing an immunity in the form of a covenant not to sue, as opposed to a license will not avoid patent exhaustion, as long as the covenant applies to sales by the covenantee.  What is not clear is whether the court, in repeatedly stating that the covenant at issue was “unconditional,” meant to leave room for a different result in the event of a “conditional” covenant not to sue.  Clearly, just as the authority to sell under a license can be limited, so can the authority to sell under a covenant not to sue, for example, by limiting the covenant to the manufacture and use of products and excluding sales.  In such a case, a sale exceeding the scope of the covenant would not be authorized, and the sale itself would infringe the patent at issue.  But it is unclear whether the Federal Circuit meant “limited” when it said “conditional.” The court’s use of terminology does resemble its confused characterization, in its LG Electronics v. Bizcom decision, of a “[license] as a sale for exhaustion purposes” and its conclusion that the patent exhaustion doctrine does not apply to a “conditional agreement.”[7]  That  decision was subsequently reversed by the Supreme Court in Quanta, albeit without expressly addressing the Federal Circuit’s findings regarding conditional sales.

Secondly, Transcore could be read to equate licenses and covenants not only for purposes of the patent exhaustion doctrine’s “authorization” requirement, but more generally.  The Federal Circuit cites to numerous cases for its seemingly general proposition that “a non-exclusive patent license is equivalent to a covenant not to sue.”  However, while it may be true that all non-exclusive licenses are covenants not to sue, it does not automatically follow that all covenants not to sue are – or are equivalent to – non-exclusive patent licenses for all purposes.  For example, while an assignee of a patent acquires the patent subject to any existing non-exclusive licenses, a covenant not to sue is commonly viewed as personal and not automatically binding on an assignee of the patent.  Moreover, it is unclear whether bankruptcy courts will extend the protections expressly afforded to “licensees” under Section 365(n) of the Bankruptcy Code to “covenantees.”  In other areas, however, courts have previously treated covenants not to sue and licenses alike.  Both have been held, for example, to constitute “authority” barring infringement for purposes of Section 271(a) of the Patent Act, and the Federal Circuit’s characterization in Transcore seems to confirm what some lower courts have previously held,[8] i.e., that the marking statute (Section 287 of the Patent Act) applies to covenants not to sue as well as licenses.

Finally, the most remarkable aspect of Transcore may well be its finding of “an implied license . . . by virtue of legal estoppel” under the circumstances of the case.  The Federal Circuit’s test for an implied license by virtue of legal estoppel is markedly different from the better-known two-prong test established in Bandag which requires, as a prerequisite for finding an implied license, that the articles sold have no non-infringing use and that the circumstances of the sale clearly indicate that a license should be implied.[9] Transcore confirms that the Bandag test is one implied license test, but by no means the only one.[10]  In fact, it could be said that the license implied under the Bandag test is a license implied in fact, while the license implied pursuant to the Transcore rule is one implied in law – regardless of the circumstances of the transaction and even despite clear language in the agreement excluding any subsequently issued patents such as the one that was then found to be licensed by implication.  Transcore suggests that it may not ever be possible to effectively exclude from a license under certain patents other patents that are necessary to practice the expressly licensed patents, except, perhaps, where it can be established that the licensee clearly understood that it would not be able to practice the licensed patents and the consideration paid was based on such understanding.

In Static Control[11] the Eastern District of Kentucky granted Static Control’s motion to reconsider a prior order that was issued before the Supreme Court’s Quanta decision. In the prior order, the court had upheld Lexmark’s single use restrictions for printer cartridges in reliance on Mallinckrodt’s conditional sale doctrine.  Upon reconsideration, the court held that Lexmark’s patent rights in its toner cartridges were exhausted as a result of their authorized sale and that Lexmark’s single use restriction was not enforceable under patent law, because “[a]fter reviewing Quanta, Mallinckrodt, and the parties’ arguments, [the court was] persuaded that Quanta overruled Mallinckrodt sub silentio.”

At issue in Static Control was Lexmark’s so called “prebate” or “Lexmark Return Program.”  Under this program Lexmark offered patented printer cartridges at two prices: a higher price for cartridges without use restrictions and a discounted price for cartridges subject to a single-use restriction and an obligation to return the used cartridge to Lexmark.  The restrictions were presented in the form of a typical shrink-wrap license agreement.  The restrictive terms were printed on the box and preceded by a statement that, by opening the package or using the cartridge, the user confirms its acceptance of the single use restriction and return obligation. Static Control supplied used toner cartridges to remanufacturers who refilled and resold them. Lexmark asserted claims of direct patent infringement and inducement of patent infringement against Static Control.

In its original order the court rejected Static Control’s argument that Lexmark’s patent rights were exhausted as a result of the authorized sale of the cartridges.  Relying heavily on Mallinckrodt, the court found that the sales were “conditional” and thus avoided exhaustion.  Upon reconsideration in light of Quanta, the court reversed its original order.  The decision reviewed in detail the Supreme Court’s exhaustion decisions preceding Quanta and concluded that the Supreme Court has consistently held that patent holders may not invoke patent law to enforce restrictions on the post-sale use of their patented products.  The court then went on to discuss the Supreme Court’s Quanta decision itself as well as the Federal Circuit’s decision and concluded that the Supreme Court – while not expressly overruling Mallinckrodt – did so sub silentio primarily based on the fact that “the Federal Circuit relied in part on Mallinckrodt in reaching its decision in LG Electronics, Inc. v. Bizcom Electronics, Inc...., the decision the Supreme Court reversed in Quanta.”

The conclusion that Quanta overruled Mallinckrodt is certainly debatable.  In fact, most commentators were surprised that the Supreme Court in Quanta did not address Mallinckrodt or the conditional sale doctrine at all – even though the Federal Circuit’s decision below rested almost entirely on its highly questionable finding of not one, but two conditional sales, and despite the fact that numerous amici briefs submitted to the Supreme Court in the Quanta case focused exclusively on this issue.  Rather than interpret Quanta as overruling Mallinckrodt, the decision might just as well be read as omitting any discussion of Mallinckrodt as irrelevant, because even if patent exhaustion could be avoided by a conditional sale, there clearly was no conditional sale in Quanta.  In finding a conditional sale in Bizcom, the Federal Circuit had stretched the concept so far that even those generally in favor of the conditional sales doctrine were struggling to justify its application in Quanta.  In fact, in its Supreme Court brief, LG Electronics all but omitted a conditional sales argument, relying instead on its re-characterization of the case as an implied license case.  Thus, another interpretation of Quanta is that it simply was not necessary for the Court to address Mallinckrodt.  The same could be said for Static Control, where the Federal Circuit itself noted that “[n]o potential buyer was required to agree to abide by the Prebate terms before purchasing a cartridge.  Thus, sales of Lexmark’s Prebate toner cartridges were authorized and unconditional, just like sales of LGE’s patented products in Quanta.”  In other words, if LGE’s sales – like Lexmark’s sales – were, in fact, unconditional, there was no need for the Supreme Court to overrule Mallinckrodt – just as there was no need for the Static Control court to so interpret Quanta.  The fact that the court did so anyway may be evidence of a more general opposition to Mallinckrodt’s basic principle.  The court in Static Control believed that post-sale restrictions on the use of patented articles should not be enforceable under patent law, notwithstanding the possibility that they may be enforceable under contract law to the extent such restrictions are validly agreed to by the party against whom they are to be enforced.  This is consistent with the criticism Mallinckrodt has faced since the very beginning – and while the Supreme Court did not clearly take sides, it did reignite the debate.  The conditional sales doctrine is being challenged, although it remains to be seen whether other courts follow Static Control.

LG Electronics v. Hitachi[12] presents another example of a district court relying on Quanta to justify a departure from controversial Federal Circuit precedent.  The facts in this case were almost identical to the facts in Quanta.  LG Electronics asserted the same patents against Hitachi that were at issue in Quanta, and the allegedly infringing acts by Hitachi were the same as those complained about in Quanta, namely, the combination of parts sold by Intel under license from LG Electronics with other components in a way that practiced LG Electronics’ patented methods.

LG Electronics argued that this case was distinguishable from Quanta, relying primarily on the fact that Intel’s authorized sale of chipsets had not occurred in the United States.  Citing the Federal Circuit’s Jazz Photo decisions, LG Electronics argued that foreign sales could not trigger exhaustion of U.S. patents regardless of whether these sales were authorized.

The district court acknowledged that the Federal Circuit’s holding in the Jazz Photo cases would prohibit a finding of patent exhaustion based on a foreign sale, but held that such a result would be inconsistent with the Supreme Court’s Quanta decision because it would permit the patent holder to do exactly the kind of “end-run” around exhaustion that the Supreme Court disapproved of in Quanta.  Even though the Supreme Court had not specifically addressed foreign sales in its decision, the district court stated that the Court’s rationale for its decision supported the conclusion “that it meant ‘authorized sales’ to include ‘authorized foreign sales.’”  Because it was undisputed that Intel’s foreign sales to Hitachi were specifically authorized by virtue of Intel’s worldwide license, the court held that these sales triggered exhaustion of LG Electronics’ patents essentially embodied in the Intel products sold.

Prior to Quanta, the Federal Circuit’s categorical ruling in Jazz Photo that only authorized sales in the United States trigger exhaustion of a United States patent had frequently been criticized.  Especially in cases where the seller had a worldwide license, the result – exhaustion if the product was first sold in the United States, but no exhaustion if the product was first sold abroad and then resold in the United States – seemed arbitrary because a first sale in the United States would have been authorized in any event.  In Hitachi, the district court seized the opportunity to depart from the Federal Circuit’s rule by broadly interpreting Quanta, despite the fact that Quanta did not expressly address the issue of foreign sales.

Practical Implications

The practical implications of the Federal Circuit’s Transcore decision could be significant.  Partly to avoid patent exhaustion, it has become common practice for many patent owners to grant limited personal covenants not to sue rather than licenses – often expressly retaining the right to assert their patents against downstream customers of the covenantee.  The Federal Circuit has now expressly confirmed what Quanta and other cases before it already implied: that for purposes of patent exhaustion a covenant not to sue amounts to authorization just as a license does.  As Quanta made clear, patent owners seeking to avoid exhaustion must limit the scope of the licensee’s or covenantee’s authorization to sell.  Sales exceeding the scope of the authorization remain unauthorized, and are themselves infringing, and thus do not trigger exhaustion.  What courts will not allow, however, is the grant of patent immunity to a party upstream while a patent owner retains the ability to assert its patents against downstream customers.

More generally, patent owners and licensees or covenantees will need to consider carefully the implications of granting or obtaining a covenant not to sue instead of a license.  While a covenant may be treated like a license for many purposes – including, for example, with respect to patent marking – it may still afford less protection than a license in the event of bankruptcy or a transfer of the underlying patent to a third party who does not agree to be bound by the covenant.

All licensors should be aware of Transcore’s holding regarding implied licenses by virtue of estoppel.  Because such a license is, in effect, a license implied-in-law, disclaimers of implied licenses and even express exclusions of certain patents may not succeed in defeating the implication of a license as to patents that are necessary to practice the expressly licensed patents.

The impact of the Static Control and Hitachi decisions is less clear at this time. Both are district court decisions that are clearly inconsistent with pre-Quanta Federal Circuit precedent.  And while both decisions rely on Quanta to justify that departure, Quanta did not clearly address either of the issues raised in Static Control or Hitachi.  Until the Federal Circuit itself either adopts or rejects the district courts’ interpretations of Quanta, it will remain unclear to what extent Mallinckrodt and Jazz Photo remain good law.


Footnotes

[1] 128 S. Ct. 2109 (June 9, 2008).

[2] See Morrison & Foerster Intellectual Property Law Update, Supreme Court Issues Ruling on Patent Exhaustion in Quanta Computer, Inc. v. LG Electronics, Inc. (June 9, 2008).

[3] See Mallinckrodt, Inc. v. Medipart, Inc. 976 F.2d 700, 708 (Fed. Cir. 1992).

[4] See Jazz Photo Corp. v. Int’l Trade Comm’n, 264 F.3d 1094 (Fed. Cir. 2001) and Fuji Photo Film Co. Ltd. v. Jazz Photo Corp., 394 F.3d 1368 (Fed. Cir. 2005).

[5] Transcore, LP v. Electronic Transaction Consultants Corp., Case No. 2008-1430, 2009 U.S. App. LEXIS 7428 (Fed. Cir., April 8, 2009).

[6] Wang Labs., Inc. v. Mitsubishi Elecs. Am., Inc.103 F.3d 1571, 1581 (Fed. Cir. 1997).

[7] LG Elecs., Inc. v. Bizcom Elecs., Inc., 453 F.3d 1364, 1370 (Fed. Cir. 2006), rev’d 128 S. Ct. 2109 (June 9, 2008).

[8] See, e.g. In re Yarn Processing Patent Validity Litigation, 602 F. Supp. 159, 169 (W.D.N.C. 1984) (“Section 287 applies to all ‘persons’ who make or sell ‘for or under’ the authority of the patentee and thus applies . . . regardless of the particular form these authorizations may take and regardless of whether the authorizations are ‘settlement agreements,’ ‘covenants not to sue’ or ‘licenses.’”).

[9] See Bandag, Inc. v. Al Bolser’s Tire Stores, Inc., 750 F.2d 903, 924-25 (Fed. Cir. 1984).

[10] Other recent Federal Circuit decisions have also found implied licenses under theories different from both the one underlying Bandag and the one underlying TranscoreSee, e.g. Jacobs v. Nintendo of Am., Inc., 370 F.3d 1097, 1101 (Fed. Cir. 2004) and Zenith Elecs. Corp. v. PDI Communications Systems, Inc., 522 F.3d 1348 (Fed. Cir. 2008).

[11] Static Control Components, Inc. v. Lexmark Int’l, Inc., Case Nos. 5:02-571 and 5:04-84, 2009 U.S. Dist. LEXIS 29479 (E.D. Ky. March 31, 2009).

[12] LG Electronics, Inc. v. Hitachi, Ltd., Case No. C 07-6511 CW, 2009 U.S. Dist. LEXIS 20457 (N.D. Cal. March 13, 2009).

Email Disclaimer

Unsolicited e-mails and information sent to Morrison & Foerster will not be considered confidential, may be disclosed to others pursuant to our Privacy Policy, may not receive a response, and do not create an attorney-client relationship with Morrison & Foerster. If you are not already a client of Morrison & Foerster, do not include any confidential information in this message. Also, please note that our attorneys do not seek to practice law in any jurisdiction in which they are not properly authorized to do so.

©1996-2017 Morrison & Foerster LLP. All rights reserved.