MEMORANDUM OPINION AND ORDER

WILLIAM C. BRYSON, Circuit Judge.

Before the Court is a motion for summary judgment of patent invalidity filed by defendants Intuit Inc. and the Hertz Corporation (Dkt. No. 158), a motion for summary judgment of laches, filed by the Hertz Corporation (Dkt. No. 159), and a motion for summary judgment of non-infringement under the doctrine of divided infringement filed by defendants Intuit Inc. and the Hertz Corporation (Dkt. No. 160). For the reasons set forth below, all three motions for summary judgment are DENIED.

I. Invalidity

A. Background

TQP Development, LLC, sued Intuit and Hertz for infringement of U.S. Patent No. 5,412,730 ("the '730 patent"). TQP asserts that the defendants' accused websites infringe the '730 patent based on their use of the SSL and/or TLS internet security protocols in combination with the RC4 encryption algorithm. The defendants contend that the '730 patent is invalid because it is anticipated by the Notes product (popularly known as Lotus Notes) that was developed by Iris Associates ("Iris") and sold to Lotus Development Corporation ("Lotus") in the late 1980s. The defendants present two theories of anticipation. First, they argue that the asserted claims of the '730 patent are invalid based on the "on-sale bar" of 35 U.S.C. § 102(b). In particular, they contend that the Notes product, implemented with the RC4 algorithm, was the subject of a sale or offer for sale more than a year before the filing date of the application that led to the '730 patent. Second, they argue that the asserted claims of the '730 patent are invalid under the "prior-invention rule" of 35 U.S.C. § 102(g), because the invention claimed in the '730 patent was first invented by another person, who did not abandon, suppress, or conceal it.

The parties do not dispute that the Notes product incorporating RC4 meets all the elements of the asserted independent claim, as required for anticipation under section 102. They disagree, however, about (1) whether Notes incorporating RC4 was the subject of a commercial sale or offer for sale more than one year before the filing date of the patent application for purposes of section 102(b); (2) whether Notes incorporating RC4 was "ready for patenting" more than one year before the filing date of the patent application, for purposes of section 102(b); and (3) whether the prior invention of Notes incorporating RC4 was suppressed or concealed within the meaning of section 102(g). Because the Court determines that the defendants have not met their burden on summary judgment to establish invalidity under section 102, it is not necessary to further inquire into whether Notes with RC4 meets the limitations of the asserted dependent claims.

B. Section 102(b): The On-Sale Bar

The "on-sale bar" rule is set forth in the portion of section 102(b) that provides that a person shall not be entitled to a patent if the invention was "on sale in this country, more than one year prior to the date of the application for patent in the United States." 35 U.S.C. § 102(b) (2006).1 The Supreme Court in Pfaff v. Wells Electronics, Inc., 525 U.S. 55, 67-68 (1998), established a two-part test for determining when the on-sale bar of section 102(b) operates to invalidate a patent. First, the product whose sale is claimed to be invalidating must have been "the subject of a commercial offer for sale" more than one year prior to the date of the patent application. Id. at 67. In this regard, the date one year prior to the patent application is referred to as the "critical date" for purposes of assessing the validity of the patent. See id. at 57.

Second, the invention embodied in the invalidating product must have been "ready for patenting" before the critical date. See Pfaff, 525 U.S. at 67. A party can satisfy the second part of the Pfaff test with "proof of reduction to practice before the critical date" or with proof that before the critical date the inventor of the invalidating product "had prepared drawings or other descriptions of the invention that were sufficiently specific to enable a person skilled in the art to practice the invention." Id. at 67-68. Although an invention need not be ready for patenting before the commercial offer for sale, there can be no offer for sale of an invention "until such time as the invention is conceived." August Tech. Corp. v. Camtek, Ltd., 655 F.3d 1278, 1289 (Fed. Cir. 2011). In order for a patent claim to be anticipated under the on-sale bar, "each and every limitation" of the claim must be found "either expressly or inherently in the device or process that was sold." Minton v. Nat'l Ass'n of Sec. Dealers, Inc., 336 F.3d 1373, 1376 (Fed. Cir. 2003) (internal quotation marks omitted).

1. Whether the Prior Invention Was On Sale Prior to the Critical Date

The application leading to the '730 patent was a continuation in part of an application filed on October 6, 1989. Therefore, even though the defendants do not concede that the '730 patent is ultimately entitled to that application's critical date, the parties agree that October 6, 1988, is the earliest possible critical date and should be deemed the correct critical date for purposes of the present motion.

The defendants assert that the source code for the version of Notes that incorporated RC4 was sold by Iris to Lotus before that critical date. The sale, they contend, occurred as early as December 7, 1984, when Iris and Lotus entered into a "Development and Option Agreement." At the time of that agreement, Iris was developing the source code for Notes. The agreement granted Lotus the option to buy the Notes source code from Iris. See Dkt. No. 158-9. According to the defendants, Lotus exercised its option on January 6, 1988. As a result, the Notes source code was delivered to Lotus no later than the spring of 1988. The defendants contend that the source code that was delivered at that time included the code for the RC4 encryption feature.

In response, TQP argues that the alleged January 1988 sale could not have been a sale of an invalidating product because RC4 had not yet been implemented in Notes at that time. TQP asserts that Iris did not have possession of RC4 in January 1988 and that Iris never had rights to sell the RC4 source code, which was developed by a third party. TQP points to the lack of any contracts between Iris and Lotus that mention the transfer of RC4 or its source code. According to TQP, the only evidence that establishes a sale of Notes incorporating RC4 is the evidence of the first commercial sales that occurred in December 1989 upon the release of Lotus Notes to the public.

The record reveals that there is a genuine dispute of fact with regard to whether Notes with RC4 was the subject of a sale or offer for sale before the critical date. The Development and Option Agreement granted Lotus the right to purchase the source code "for the [Notes] Software developed by Iris" (emphasis added) pursuant to the agreement. See Dkt. No. 158-9, at 3, 11. RC4, however, was not "developed by Iris"; it was developed by Dr. Ron Rivest at his company, RSA Data Security, Inc. Dkt. No. 158-10, at 3-4. Iris employee Alan Eldridge, who requested that Dr. Rivest develop the algorithm that became RC4 and who subsequently incorporated Dr. Rivest's algorithm into Notes, testified in an earlier case brought by TQP that he incorporated the source code he received from RSA without making any changes to it. Dkt. No. 158-11, at 8-9. Mr. Eldridge also testified that Iris did not own the RC4 code and that RSA retained all the rights to that code. Dkt. No. 171-4, at 4. In fact, Mr. Eldridge testified that he was the only person at Iris who was allowed to see the source code. Dkt. No. 158-11, at 8. Evidence adduced by TQP tends to show that Lotus's right to use RC4 in Notes stemmed from a licensing agreement between Lotus and RSA, not from the sale of the RC4 source code as a component of the Notes product developed by Iris. Dkt. No. 171-4, at 4.

TQP has pointed to evidence that whatever was sold to Lotus did not include the rights to the RC4 source code, even if Iris had integrated the RC4 source code into Notes with the expectation that Lotus would obtain a separate license from RSA for use of the RC4 code. Because there is evidence indicating that only the Notes source code—not the embedded RC4 source code—was the subject of the disputed sale from Iris to Lotus, a reasonable jury could find that the sale was not the sale of a product that contained each element of the disputed claims of the '730 patent. See Scaltech Inc. v. Retec/Tetra, L.L.C., 178 F.3d 1378, 1383 (Fed. Cir. 1999) (the invalidating invention that was offered for sale must meet each limitation of the disputed patent claim).2 For that reason, the defendants are not entitled to summary judgment on the "onsale bar" issue.

2. Whether the Prior Invention Was Ready for Patenting Before the Critical Date

Unlike the question whether the invention was on sale before the critical date, there is no genuine dispute of material fact as to whether the Notes product, incorporating RC4, was ready for patenting prior to the critical date.

As an initial matter, the evidence establishes that RC4 was ready for patenting prior to the critical date. In April 1988, Dr. Rivest submitted the RC4 source code to the National Security Agency for export authorization associated with the Notes product. See Dkt. No. 158-20, at 3-5. Dr. Rivest testified that he "would not have submitted an incomplete or untested algorithm to the NSA for review" and that "[a]t the latest, the RC4 algorithm was complete and able to successfully perform encryption and decryption" by April 1988. Id. at 4; Dkt. No. 158-10 at 4.3 Additionally, the evidence establishes that Dr. Rivest's RC4 code was not substantially altered when it was integrated with the Notes code. Dkt. No. 158-11, at 8-9; Dkt. No. 158-7, at 20.

As for the combination of Notes and RC4, Iris founder Raymond Ozzie states that RC4 was integrated with Notes no later than the spring of 1988. Dkt. No. 158-7, at 7, 10-11. Alan Eldridge, the Iris employee who was responsible for integrating the RC4 source code into Notes, states that the RC4 code was incorporated into Notes in February 1988. Dkt. No. 158-11, at 6. The evidence also shows that Notes beta testers were using a version of Notes that included RC4 by mid-1988. Dkt. No. 158-7, at 14.

TQP asserts that the evidence that RC4 was incorporated into Notes by the spring of 1988 is insufficient because the Notes product was still in a developmental and beta-testing phase at that time. Therefore, according to TQP, the version of Notes that was in existence in the spring of 1988 was not a final product. TQP's argument, however, misconstrues the meaning of "ready for patenting." An invention that has been reduced to practice is ready for patenting. Pfaff, 525 U.S. at 67. In order for an invention to be reduced to practice, it does not need to be implemented in a final commercial product. In fact, reduction to practice is an event that often precedes the creation of a commercial product and frequently occurs long before the commercial embodiment of the invention is developed, tested, and ready for marketing. See, e.g., Flex-Rest, LLC v. Steelcase, Inc., 455 F.3d 1351, 1359 (Fed. Cir. 2006) (more than six-month period between reduction to practice and commercialization as the result of "reasonable steps to bring the invention to market"); Dow Chem. Co. v. Astro-Valcour, Inc., 267 F.3d 1334, 1343 (Fed. Cir. 2001) (30-month period between reduction to practice and commercialization during which prior inventor procured financing and addressed safety considerations). Therefore, evidence that Notes was not in its final commercial form at the time RC4 was integrated with it in the spring of 1988 does not by itself create a genuine issue of material fact regarding whether the invention that would anticipate the '730 patent was ready for patenting at that time.

TQP has failed to point to any evidence that would tend to establish that the version of Notes incorporating RC4 that was in existence in the spring of 1988 did not embody the invention disclosed in the '730 patent. The Court therefore concludes that there is no genuine dispute of material fact regarding whether the invalidating invention embodied by Notes incorporating RC4 was ready for patenting before the critical date.

C. Section 102(g): Invention Previously Made in the United States

A patent claim is also invalid due to anticipation if the invention disclosed in the claim "was made in this country by another inventor who had not abandoned, suppressed, or concealed it." 35 U.S.C. § 102(g) (2006).

In order to prove that the invention was made by a prior inventor, the party challenging validity must show by clear and convincing evidence that the prior inventor either was the first to reduce the invention to practice or was the first to conceive of the invention and then exercised reasonable diligence in reducing it to practice. Fox Grp., Inc. v. Cree, Inc., 700 F.3d 1300, 1304 (Fed. Cir. 2012). Once the party challenging the patent's validity "has proven by clear and convincing evidence that the `invention was made in this country by another inventor,' the burden of production shifts to the patentee to produce evidence sufficient to create a genuine issue of material fact as to whether the prior inventor has suppressed or concealed the invention." Apotex USA, Inc. v. Merck & Co., 254 F.3d 1031, 1037 (Fed. Cir. 2001) (citation omitted).

There are two ways a patentee can show that a prior inventor has suppressed or concealed an invention. Fox Grp., 700 F.3d at 1305. The more direct way is to show that the prior inventor actively and intentionally suppressed or concealed the invention. Id. "Intentional suppression occurs when an inventor `designedly, and with the view of applying it indefinitely and exclusively for his own profit, withholds his invention from the public.'" Flex-Rest, LLC v. Steelcase, Inc., 455 F.3d 1351, 1358 (Fed. Cir. 2006), quoting Paulik v. Rizkalla, 760 F.2d 1270, 1273 (Fed. Cir. 1985) (en banc); Fujikawa v. Wattanasin, 93 F.3d 1559, 1567 (Fed. Cir. 1996) ("Intentional suppression . . . requires more than the passage of time. It requires evidence that the inventor intentionally delayed filing [for a patent] in order to prolong the period during which the invention is maintained in secret."). The second way a patentee can show that a prior inventor suppressed or concealed the invention is to offer evidence of an unreasonable delay in making the invention publicly known, which can lead to an inference of suppression or concealment. Fox Grp., 700 F.3d at 1305; Apotex, 254 F.3d at 1038.

Once the patentee has met its burden of production, "the party alleging invalidity under § 102(g) must rebut any alleged suppression or concealment with clear and convincing evidence to the contrary." Apotex, 254 F.3d at 1038. Delay in publicly disclosing an invention may be excused "upon proof that the first inventor engaged in reasonable efforts to bring the invention to market," Checkpoint Sys., Inc. v. U.S. Int'l Trade Comm'n, 54 F.3d 756, 762 (Fed. Cir. 1995), or proof that the delay was otherwise reasonable, see Fujikawa, 93 F.3d at 1568. The party challenging the patent bears the ultimate burden of proof in establishing invalidity. Apotex, 254 F.3d at 1037-38.

1. Whether the Prior Invention Was Suppressed or Concealed

TQP has not substantively disputed that the invention disclosed in the '730 patent was first made in this country by another.4 See 35 U.S.C. § 102(g). Instead, the parties disagree about whether the prior invention was suppressed or concealed from the public.

The defendants argue that the prior invention was not suppressed or concealed because Iris and Lotus were "engaged in activities to commercialize the Notes product from at least the spring of 1988 until its first copy was sold to the general public in December of 1989." Those activities included demonstrations to and discussions with Microsoft and beta testing of Notes with application partners such as Reuters. Dkt. No. 158-7, at 10-11,15, 18. The discussions with Microsoft, which included some discussion of Notes' encryption features, were not even secret, according to the defendants, because Microsoft was not subject to a nondisclosure agreement. Id. at 11-12. The defendants further assert that the use of RC4 in Notes was not secret, and that the trade-secret status of the "human-readable source code for the particulars of the mathematical algorithm used in RC4" is irrelevant because public knowledge of the use of RC4 is enough to show that it was not suppressed or concealed. The defendants further assert that Microsoft or Reuters, which were provided with the Notes product prior to its release, could have reverse engineered the details of RC4, even though there is no evidence that they actually did so.

TQP has offered evidence that RC4 was kept as a trade secret by Dr. Rivest and RCA, and that it was publicly disclosed only after a hacker reverse engineered the algorithm and posted the results to a website in 1994. See Dkt. No. 171-4, at 8, 17. TQP has also shown that the relationships between Iris, Lotus, and RSA were the subjects of multiple confidentiality agreements. See Dkt. No. 171-5. For example, the licensing agreement between Lotus and RSA for the use of RC4 in Notes provided that Lotus would keep the source code confidential, id. at 9, as did a nondisclosure agreement between Iris and RSA, id. at 13. Indeed, Mr. Eldridge testified that he was the only person at Iris who was allowed to see the RC4 code, Dkt. No. 158-11, at 8, and even he was allowed to do so only after agreeing not to obtain employment implementing "crypto layer" code for 20 years, Dkt. No. 171-4, at 5.

With respect to the disclosures made to Microsoft, which was not subject to a nondisclosure agreement, the evidence does not clearly establish that the details of RC4 relevant to the '730 patent were disclosed. Instead, Raymond Ozzie, the founder of Iris, testified that the discussions between Iris and Microsoft about encryption did not concern the algorithm itself, but concerned the use of the algorithm in Notes. See Dkt. No. 158-7, at 12. Mr. Ozzie testified that Iris probably told Microsoft that it was using a "really fast stream cipher or something like that" in Notes. Id. Drawing all inferences in TQP's favor, that testimony does not establish that the portions of RC4 relating to the claims of the '730 patent were discussed with Microsoft.

With respect to the trade secret status of RC4, the defendants have offered some evidence that the secret part of RC4 was not the algorithm itself, but the software techniques that allowed the algorithm to run quickly. Dkt. No. 158-11, at 8, 10 (testimony of Mr. Eldridge). However, their evidence does not establish that the aspects of RC4 relevant to the '730 patent's claims were not part of RCA's trade secret during the relevant time. Furthermore, TQP has pointed to testimony tending to establish that "the algorithm" for RC4 was never officially released by RSA. Dkt. No. 171-4, at 4. Indeed, Mr. Eldridge himself testified that he had "no evidence one way or the other" as to whether RSA ever officially released the RC4 algorithm. See id.

The defendants are correct that, absent an unreasonable delay, there can be no suppression or concealment by a prior inventor "if the prior inventor takes affirmative steps to make the invention publicly known." They add, however, that in taking those affirmative steps, the prior inventor makes the invention publicly known even if he does not disclose the "inner workings" of the invention to the public. To support that legal proposition, the defendants cite a 1987 district court case from the District of Delaware, Friction Div. Prods., Inc. v. E.I. DuPont de Nemours & Co., 658 F.Supp. 998, 1013-14 (D. Del. 1987), aff'd, 883 F.2d 1027 (Fed. Cir. 1989) (non-precedential decision). The Federal Circuit, however, has not drawn that distinction. The circuit court has made clear that a finding of suppression and concealment requires evidence of the inventor's unreasonable delay in making "the invention" publicly known. See Fox Grp., 700 F.3d at 1305. When the "inner workings" are the essence of the invention, it is those "inner workings" that must not be suppressed or concealed in order for the invention to be prior art under section 102(g).

That principle is consistent with the Federal Circuit's decision in Lockwood v. American Airlines, Inc., 107 F.3d 1565 (Fed. Cir. 1997), on which the defendants rely to support their argument that the details of the RC4 code are irrelevant. In Lockwood, the Federal Circuit found that American Airlines' SABRE reservation system was prior art that could be used to invalidate the patent at issue in the case. The court rejected the patentee's argument that one skilled in the art would not be able to build and practice the prior invention without access to the secret aspects of the SABRE system, because "public use of the high-level aspects of the SABRE system was enough to place the claimed features of the [disputed] patent in the public's possession." Id. at 1570 (emphasis in original). The court explained that the patentee could not rebut that point with evidence "that other, unclaimed aspects of the SABRE system," i.e., the inner workings, were not publicly available. Id. (emphasis in original). See also Dey L.P. v. Sunovion Pharm., Inc., 715 F.3d 1351, 1359 (Fed. Cir. 2013) (reversing summary judgment of invalidity under the public-use prong of section 102(b) because "a reasonable jury could conclude that if members of the public are not informed of, and cannot readily discern, the claimed features of the invention in the allegedly invalidating prior art, the public has not been put in possession of those features").5

In summary, the defendants have failed to show that there is no disputed issue of material fact as to whether the prior invalidating invention was abandoned, suppressed, or concealed under section 102(g). They have also failed to establish that the prior invention was "on sale" within the meaning of 102(b). The defendants' motion for summary judgment of invalidity of the '730 patent is therefore denied.

II. Laches

A. Background

The '730 patent issued to inventor Michael Jones on May 2, 1995, and expired on May 2, 2012. After it issued, Mr. Jones assigned the patent to his employer, Telequip Corporation. Dkt. No. 170-3. At least while Mr. Jones was employed by Telequip, the company was not actively involved in seeking out infringers of the '730 patent. Dkt. No. 174-1, at 11-12. The cryptography division of Telequip—referred to as the "Secure Coprocessing Division"— was shut down by 1997 so that Telequip could focus on its coin-dispensing business. Dkt. No. 170-4, at 3; Dkt. No. 170-5, at 4; Dkt. No. 170-6, at 3. When that division was shut down, Mr. Jones discontinued his employment with Telequip so that he could start another company, although he remained a corporate director and shareholder of Telequip. Dkt. No. 170-6, at 4. In 2006, Telequip's assets, including the '730 patent, were sold to Crane Corporation. Dkt. No. 170-6, at 4. Plaintiff TQP purchased the '730 patent from Crane in 2008. Id. at 6-7. TQP began filing infringement suits asserting the '730 patent in 2008, although it did not sue Hertz until November 2012. The accused Hertz websites became publicly available starting in May 1998, with the last accused website coming online in March 2011.

Hertz asserts that TQP is barred from recovering pre-suit damages under the equitable doctrine of laches because the owners of the '730 patent unreasonably delayed asserting infringement, thereby prejudicing Hertz. According to Hertz, there is no genuine issue of material fact bearing on the availability of the defense of laches. Moreover, Hertz points out that because the suit against Hertz was not filed until after the '730 patent expired, laches would be a complete defense for Hertz.

B. Legal Principles

Laches is an equitable defense that precludes a plaintiff from seeking presuit damages based on his unreasonable delay in bringing suit. See A.C. Aukerman Co. v. R.L. Chaides Constr. Co., 960 F.2d 1020, 1031, 1041 (Fed. Cir. 1992) (en banc). To invoke the defense of laches, the defendant must prove (1) that the "plaintiff delayed filing suit for an unreasonable and inexcusable length of time from the time the plaintiff knew or reasonably should have known of its claim against the defendant"; and (2) "the delay operated to the prejudice or injury of the defendant." Id. at 1032. There is a presumption that those elements of laches are established "upon proof that the patentee delayed filing suit for more than six years after actual or constructive knowledge of the defendant's alleged infringing activity." Id. at 1035-36. Once the presumption applies, the burden of production shifts to the plaintiff, who must introduce rebuttal evidence showing "an excuse for the delay or that the delay was reasonable" or showing that the defendant suffered neither defense nor economic prejudice. Id. at 1038. The ultimate burden of persuasion to establish the affirmative defense of laches, however, rests with the defendant. Id. Even when the defense of laches is established, it does not automatically bar the plaintiff's suit. Rather, "[t]he application of the laches defense is discretionary, and as an equitable matter, the district court is to look to all the facts and circumstances of the case and weigh the equities of the parties." Gasser Chair Co., Inc. v. Infanti Chair Mfg. Corp., 60 F.3d 770, 773 (Fed. Cir. 1995).

The period of delay in bringing suit begins at the time the patentee or his predecessor in interest "has actual or constructive knowledge of the defendant's potentially infringing activities." Wanlass v. Gen. Elec. Co., 148 F.3d 1334, 1337 (Fed. Cir. 1998); see Eastman Kodak Co. v. Goodyear Tire & Rubber Co., 114 F.3d 1547, 1559 (Fed. Cir. 1997) ("A patentee cannot avoid the consequences of his laches by transferring the patent."), abrogated on other grounds, Cybor Corp. v. FAS Techs., Inc., 138 F.3d 1448, 1455 (Fed. Cir. 1998); I/P Engine, Inc. v. AOL Inc., 915 F.Supp.2d 736, 743 (E.D. Va. 2012) ("[I]t is settled law in the United States `that in determining the length of delay, a transferee of the patent must accept the consequences of the dilatory conduct of immediate and remote transferors.'").

Because constructive knowledge triggers the start of the laches clock, patentees have a duty to police their rights. Gen. Elec., 148 F.3d at 1338. The extent of that duty, and thus the question whether a patentee is deemed to have constructive knowledge of alleged infringing activity, is a question of fact.6 See Wanlass v. Fedders Corp., 145 F.3d 1461, 1464 (Fed. Cir. 1998) (vacating summary judgment of laches because "the facts as developed for the summary judgment motion are not without genuine disputes on material issues as to whether [the plaintiff] knew or reasonably should have known of [the defendant's] allegedly infringing activity before the critical date" for laches); Advanced Cardiovascular, Inc. v. SciMed Life Sys., Inc., 988 F.2d 1157, 1162-63 (Fed. Cir. 1993) ("Whether and when" the duty of inquiry is placed upon one against whom laches is asserted is a "material question of fact."); Emhart Indus., Inc. v. Universal Instruments Corp., 1992 WL 442248, at *6 (N.D.N.Y. June 22, 1992) ("Whether . . . plaintiffs possessed knowledge of sufficient facts so as to create the duty of inquiry, is a question of fact precluding the grant of either the presumption or summary judgment."); see also Robertson v. Seidman & Seidman, 609 F.2d 583, 591 (2d Cir. 1979) (stating with respect to whether a fraud claim was time barred that issues of the plaintiff's "due diligence and constructive knowledge depend on inferences drawn from the facts of each particular case"); Lenz v. Associated Inns & Rests. Co. of Am., 833 F.Supp. 362, 371 (S.D.N.Y. 1993) ("[T]he question of whether a plaintiff exercised reasonable diligence is usually a question of fact for the jury to decide.").

Circumstances that give rise to constructive knowledge include "`pervasive, open, and notorious' activities that a reasonable patentee would suspect were infringing" such as "sales, marketing, publication, or public use of a product similar to or embodying technology similar to the patented invention, or published descriptions of the defendant's potentially infringing activities" where those activities are "sufficiently prevalent in the inventor's field of endeavor." Gen. Elec., 148 F.3d at 1338. If a patentee's ignorance of infringement is justifiable, however, constructive knowledge will not be imputed to him. Id.

C. Whether Hertz has Shown Constructive Knowledge of the Alleged Infringement

Hertz argues that it is entitled to the presumption of laches and that TQP has insufficient evidence to rebut the presumption. Hertz does not argue that TQP's predecessors in interest had actual knowledge of Hertz's alleged infringement. Rather, Hertz contends that TQP's predecessors in interest had constructive knowledge of the alleged infringement more than six years before TQP filed suit. Hertz asserts that the alleged infringing acts were open and notorious before the November 2, 2006, laches critical date (six years before suit was filed) because "Hertz began using encryption on its publicly available, accused websites" in 1998. Hertz points to evidence that a member of the public accessing Hertz's accused websites would have been able to tell that the websites used the SSL protocol, as indicated by the use of the "https" prefix in the accused website address. The SSL protocol is part of the allegedly infringing combination of SSL and RC4. See Dkt. No. 159-3.

Hertz notes that the RC4 algorithm was publicly available as early as 1994 when it was disclosed by a hacker on the Cypherpunk Newsgroup. See Dkt. No. 159-6. Moreover, Hertz argues that TQP relied on publicly available information for evidence of infringement and that the prior owners of the '730 patent could and should have done the same. In addition, Hertz points to the deposition testimony of the inventor, who testified that while he was president of Telequip, the original assignee of the '730 patent, the company was not investigating potential infringement or attempting to enforce the '730 patent. Therefore, according to Hertz, the prior owners of the '730 patent unreasonably slept on their rights—conduct that is chargeable to TQP.

As the party moving for summary judgment on an issue on which it bears the burden of proof, Hertz is required to come forward with evidence demonstrating that there is no genuine issue of material fact and that it is entitled to judgment as a matter of law. Fed. R. Civ. P. 56(a); Celotex Corp. v. Catrett, 477 U.S. 317, 322-23 (1986); Meyers v. Brooks Shoe Inc., 912 F.2d 1459, 1461 (Fed. Cir. 1990), overruled on other grounds, A.C. Aukerman Co. v. R.L. Chaides Constr. Co., 960 F.2d 1020, 1038-39 (Fed. Cir. 1992) (en banc). After considering the evidence offered by both parties, the court is required to enter summary judgment for the moving party only if "there can be but one reasonable conclusion" as to the outcome. Anderson v. Liberty Lobby, Inc. 477 U.S. 242, 250 (1986). "If reasonable minds could differ as to the import of the evidence," then the movant has not met its burden. Id. at 250-51.

The evidence before the Court at this juncture is insufficient to establish as a matter of law that the prior owners of the '730 patent had constructive knowledge of Hertz's alleged infringement prior to the laches critical date. Although the material underlying facts of this case are not in dispute, there remains a dispute as to the ultimate issue of fact: whether the owners of the '730 patent should have known of Hertz's allegedly infringing activity. The facts that Hertz's publicly accessible websites used SSL beginning in 1998 and that source code disclosing the RC4 algorithm had been posted on a hacker site in 1994, do not establish that a reasonably diligent patentee would have had reason to know that the use of SSL in combination with RC4 was proliferating on the internet, much less that Hertz was taking advantage of that combination before the laches critical date.

In order to reach that conclusion, the Court would have to draw several inferences in Hertz's favor. For example, the Court would have to infer that a reasonable patentee would have recognized the aspects of RC4 relevant to the '730 patent based on the computer code published on the Cypherpunk Newsgroup. The patentee would then have to recognize that the combination of RC4 and SSL might infringe the '730 patent, even though neither alone does. The Court would then be required to infer that a reasonable patentee would have known that the alleged infringing combination of RC4 and SSL was being used by internet websites, and in particular that Hertz had a website using that combination. It would be impermissible for the Court on summary judgment to draw those inferences in favor of Hertz, when the opposite inferences may reasonably be drawn in favor of TQP, the nonmovant. See Anderson, 477 U.S. at 255. The mere fact that the accused websites were publicly available does not by itself establish as a matter of law that a reasonable patentee would have known that the encryption programs used on those websites might be infringing. For that reason, Hertz's evidence does not establish that there is no genuine issue of material fact left for trial with regard to whether the owners of the '730 patent had constructive knowledge of Hertz's alleged infringement prior to November 2, 2006.

Hertz argues that "it is important to note that the SSL/RC4 is a standardized protocol and encryption cipher suite that was well-known and widely used going back to at least the late 1990s." That attorney argument, however, is not evidence and thereby does not assist TQP in meeting its burden of establishing that there is no genuine issue of material fact with respect to constructive knowledge.

The Court's decision to deny summary judgment is supported by the Federal Circuit's decisions. The two primary Federal Circuit cases bearing on the issue whether a patentee has constructive knowledge of infringement were decided by the same panel and concerned the same patent and the same infringing technology. In Wanlass v. General Electric Co., 148 F.3d 1334 (Fed. Cir. 1998), the Federal Circuit affirmed summary judgment of laches, agreeing with the district court that the patentee had constructive knowledge of the defendant's alleged infringement before the laches critical date. Id. at 1339. In a different case involving the same patentee and decided ten days later, Wanlass v. Fedders Corp., 145 F.3d 1461 (Fed. Cir. 1998), the Federal Circuit vacated a summary judgment of laches, holding that the record on summary judgment did not establish that the plaintiff had constructive knowledge of the alleged infringement as a matter of law.

In both General Electric and Fedders, the patentee needed to test suspected air conditioning motors in order to determine whether the motors infringed the patent in suit. In General Electric the court of appeals held that it was unreasonable for the patentee to wait more than 10 years to test a General Electric product for infringement after initial tests in the late 1970s revealed no infringement and General Electric had indicated in communications with the patentee that it thought the patent was invalid. 148 F.3d at 1339-40. The court rejected the plaintiff's argument that it would have been too burdensome to continue testing General Electric products after the initial test, even though between 800 and 900 General Electric products potentially contained the allegedly infringing technology. Id. The court concluded that if the patentee had been reasonably diligent about enforcing his rights by testing General Electric products at some point during the pre-filing period of more than ten years, he would have discovered General Electric's alleged infringement. Id. at 1340.

In Fedders, by contrast, the court held that the same patentee's duty to investigate infringement did not extend to the defendant's products. 145 F.3d at 1465-66. The distinction between the cases was that in General Electric the defendant had put the patentee "on notice that [it] was a potential infringer" by way of a prior communication with the patentee in which the defendant indicated that it believed the patent was invalid. See Fedders, 145 F.3d at 1465 n.3. In Fedders, however, there was "virtually no evidence of communication" between the patentee and alleged infringer, and "insufficient evidence . . . on the summary judgment record to show [the patentee] had information suggesting that [the defendant] was a likely infringer." Id.

The summary judgment record in this case is more like the record in Fedders than that in General Electric. In this case there is no evidence of any communication between the '730 patent owners and Hertz that would have put the owners of the '730 patent on notice of Hertz's alleged infringement. As in Fedders, Hertz has not established on the summary judgment record that the patent owners had any particular reason to test the Hertz website prior to November 2, 2006. See Fedders, 145 F.3d at 1465. Although it is true that the prior owners of the '730 patent, Telequip and Crane, "could not simply ignore any and all evidence of potentially infringing activity" even if they were not active in the industry, the patentee's duty to investigate a particular infringing product does not arise until "publicly available information about it should have led [the patentee] to suspect that product of infringing." Id. at 1466. For the reasons already stated, Hertz failed to establish that there is no genuine issue of material fact as to whether the publicly available information about SSL, RC4, the combination of SSL and RC4, and Hertz's use of that technology should have led the prior owners of the '730 patent to suspect Hertz of infringing.

For the above reasons, Hertz's motion for summary judgment of laches is denied.

III. Divided Infringement

A. Background

The defendants have moved for summary judgment of noninfringement based on the doctrine of divided infringement. They contend that TQP's claim of direct infringement fails because the defendants only performed certain steps of the claimed method. The other steps, according to the defendants, were performed by clients who were acting independently of the defendants, and not under their direction and control. Because no single actor performed all of the limitations of the claims of the '730 patent, the defendants argue that there is "divided infringement," and no single party can be held liable for direct infringement. According to the defendants' theory, the defendants perform the steps involving the generation and transmission of the encrypted data, while the clients perform the steps involving the receipt and decryption of the data.

B. Legal Principles

Direct infringement requires proof that the defendant "has practiced each and every element of the claimed invention." BMC Res., Inc. v. Paymentech, L.P., 498 F.3d 1373, 1380 (Fed. Cir. 2007). In the case of method claims, the Federal Circuit has rejected claims of liability for infringement in cases in which "several parties have collectively committed the acts necessary to constitute direct infringement, but no single party has committed all of the required acts." Akamai Techs., Inc. v. Limelight Networks, Inc., 692 F.3d 1301, 1307 (Fed. Cir. 2012) (en banc), rev'd on other grounds, 134 S.Ct. 2111 (2014).

That principle is subject to an exception for cases in which the second party, who performs some of the steps of a claimed method, is acting under the direction or control of the first party, who is charged with infringement. In that situation, every step is attributable to the controlling party. Muniauction, Inc. v. Thomson Corp., 532 F.3d 1318, 1329 (Fed. Cir. 2008). "[M]ere `arms-length cooperation' will not give rise to direct infringement by any party." Id. Direction or control requires more than encouragement or suggestions by the first actor to the second. See Emtel, Inc. v. Lipidlabs, Inc., 583 F.Supp.2d 811, 839 (S.D. Tex. 2008) ("Making information available to the third party, prompting the third party, instructing the third party, or facilitating or arranging for the third-party's involvement in the alleged infringement is not sufficient."). If the second actor is free, either legally or factually, to disregard the suggestions or encouragement from the first actor, the first actor is not considered to direct or control the actions of the second. See Aristocrat Techs. Austl. Pty Ltd. v. Int'l Game Tech., 709 F.3d 1348, 1362 (Fed. Cir. 2013).

C. Whether the Defendants Direct or Control All the Steps of the Claim

It is undisputed that the defendants' servers do not perform all the steps of the claimed method. The defendants' server performs the steps of encryption and transmission, while the remaining steps are performed by clients' computers, with which the defendants' servers are in communication. The argument over divided infringement therefore comes down to whether, when the defendants practice the steps of the claimed invention involving transmitting encrypted data, they direct or control the steps that are performed by the receiving client computers. The Court concludes that there is a genuine factual dispute as to that question and that summary judgment therefore cannot be granted on the divided infringement issue.

The defendants place heavy reliance on the fact that their clients act independently in various ways in connection with the transmission and receipt of encrypted data. For example, the defendants point out that the clients "may choose not to visit the accused websites" and "the client browsers may be configured to connect to Defendants' accused websites without using the accused RC4 encryption algorithm." While those points are true, they are immaterial to the question whether the claimed method steps performed by a client computer are performed at the direction or control of the defendants' servers.

The steps of the claims begin with providing the seed value to both the transmitter and the receiver, encompass the transmission and receipt of the data, and end with the decryption of the encrypted data at the receiver. Those steps presume that the receiver (i.e., the client) and transmitter have already decided to engage in an encrypted communication. That choice by the client/receiver does not constitute the performance of any step in the patent claims, but only establish the setting within which infringement may occur. The same is true of the possibility that some client browsers will not be configured to use the RC4 encryption algorithm or that the RC4 algorithm will not be used for some reason. TQP has offered evidence that if the RC4 algorithm is enabled on the clients' browsers when the clients visit the defendants' websites, the defendants' servers will dictate that RC4 be used. Dkt. No. 168-1, at 8. If the clients' browsers do not offer RC4 as an encryption algorithm, RC4 will not be used. However, that is not to say that that the client's computer is not directed or controlled by the defendant's server; it simply establishes that in such an instance no infringement takes place. The alleged act of infringement occurs only if, and after, the RC4 algorithm is chosen as the means of encrypting the communicated data. The defendants' argument on this point is as flawed as the argument that when a physician performs a patented medical procedure on a patient that requires some action by the patient—e.g., swallowing a pill—there is "divided infringement" because the patient could have decided not to undergo the medical procedure in the first place.

In an effort to emphasize the independence of the clients' actions in the context of the accused method, the defendants assert that "the selection of the cipher suite—including the encryption algorithm—is the result of a negotiation between two independent entities." But the evidence offered by TQP, viewed in the light most favorable to it, is that the "negotiation" always results in the choice of RC4 if that algorithm is available. Dkt. No. 168-1, at 8. And in any event, the "negotiation" and selection of an encryption algorithm are not limitations of the claims; as noted, the claimed method begins after the encryption algorithm is selected.

The same is true of the defendants' observation that the clients' computers "are not provided by, programmed by, or configured by Defendants." All that conduct is prefatory to the steps recited in the claim. The client's decisions leading up to the performance of those claimed steps do not create a divided infringement scenario, because those decisions do not correspond to any claim limitations.

The key question is whether the defendants' servers "direct and control" the client computers once RC4 is selected as the encryption algorithm and the transmission process begins. TQP has offered evidence through its expert that the defendants' servers direct or control the client computers because, once the process begins, the steps taken by the servers in encrypting and transmitting data automatically produce a predictable, corresponding response in the client computers that receive and decrypt the data. See Dkt. No. 168-1, 6-10. According to TQP's evidence, the use of the RC4 algorithm in both the server and client computers dictates that the steps taken by the server at the encryption and transmission stage result in performance of the corresponding steps in the receiver. Thus, according to TQP's evidence, once the respective computers are suitably programmed and the RC4 algorithm is selected, the defendants' servers dictate the response of the client computers that will perform the "receiver stage" steps of the claimed process. In the Court's judgment, that is sufficient to create a factual question as to whether defendants' servers exercise "direction or control" over a client's computer for purposes of the doctrine of divided infringement.

The defendants cite a number of cases in support of their divided infringement argument, contending that the defendants' servers do not "direct or control" the performance of the recited steps that are performed by the clients' computers. None of those cases, however, supports the broad point for which the defendants are contending. Each of the cases involves discretionary conduct by the actors who are alleged to have been subject to the alleged infringer's direction and control. In that context, the courts have declined to find an exercise of direction or control. Thus, in Aristocrat Technologies Australia Pty Ltd. v. International Game Technology, 709 F.3d 1348 (Fed. Cir. 2013), the method claims of a gaming machine patent recited various steps performed by the gaming machine to award a progressive jackpot to a player. However, the claims also required action by the player to, for example, make a wager. Id. at 1350. The patentee argued that the alleged infringer directed or controlled players' actions in that regard by providing free game credits to induce them to gamble on the allegedly infringing machines. Although the issuance of free credits encouraged players to gamble, the court found that players were not obligated to use the free credits. Id. at 1362. The court therefore concluded that the player's actions were voluntary, and not controlled or directed by the gaming machines, which performed the other steps of the method. Id. at 1362-62.

The Federal Circuit applied similar analysis in Voter Verified, Inc. v. Premier Election Solutions, Inc., 698 F.3d 1374 (Fed. Cir. 2012). The pertinent claims at issue in that case recited automated methods for voting in an election, which provided for self-verification of the accuracy of a voter's ballot. Id. at 1377. The defendant argued that certain of the claim steps had to be performed by the voter and thus involved conduct that was not subject to the "direction or control" of the defendant. The court agreed, holding that the defendants at most "provide[d] instructions on how to use their systems." Id. at 1384. According to the claims, it was the voter who would decide "whether a printed ballot is an acceptable or unacceptable representation of his or her intended vote." Id. at 1383. The court concluded that "the fact that an accused infringer `controls access to its system and instructs [others] on its use is not sufficient to incur liability for direct infringement.'" Id. at 1384, quoting Muniauction, Inc. v. Thomson Corp., 532 F.3d 1318, 1330 (Fed. Cir. 2008).

The same is true of several other cases on which the defendants rely. See PA Advisors, LLC v. Google, Inc., 706 F.Supp.2d 739, 748 (E.D. Tex. 2010) (patent claiming a search function not infringed because one limitation required that computer user initiate the search query, which was not conduct controlled by the defendants); Deep9 Corp. v. Barnes & Noble, Inc., 2012 WL 4336726 (W.D. Wash. Sept. 21, 2012) (patent on method for updating information via a network not infringed because defendant seller of the "Nook" device did not exercise direction or control over the choices made by users to connect the device to the Internet in order to cause the "downloading" and "updating" steps of the asserted claims); Digital Impact, Inc. v. Bigfoot Interactive, Inc., 2007 WL 2729568, at *5-*7 (N.D. Cal. Sept. 19, 2007) (patent on method for sending and tracking emails not infringed by defendant because claims required that email recipient open email, which was a discretionary step not controlled by the defendant). In each of those cases, some of the steps were to be performed by parties acting volitionally and having discretion as to whether to perform the recited steps. Those cases do not apply here where, according to TQP's evidence, the response of the client computer acting as a receiver is dictated by the selection of RC4 as the encryption algorithm and the actions of the server in encrypting and transmitting data.

The defendants rely on Global Patent Holdings, LLC v. Panthers BRHC LLC, 586 F.Supp.2d 1331 (S.D. Fla. 2008). That case involved a claim reciting a method for downloading responsive data from a remote server (e.g., a website). The claim included the step of "identifying a query via a data input means and inputting said query to remote query and data retrieval means." 586 F. Supp. 2d at 1332 n.1. The remaining steps involved transmitting the query to the remote server, optionally compressing the response, receiving the response, decompressing a compressed response if needed, and displaying a presentation corresponding to the response on an output means. Id.

The court focused on the first step of identifying and inputting a query for the remote server/website. The plaintiff acknowledged that the first step was performed by a third party, but argued that it was performed pursuant to the direction or control of the defendant website operator. The court understandably held that there was no direction or control by the website operator defendant over the third-party website user. 586 F. Supp. 2d at 1335.

In the course of its ruling, the court stated that the remote users were not "contractually bound to visit the website." 586 F. Supp. 2d at 1335. The defendants seize on that language to support their contention that there can be no direction or control in this case because the clients were not required to visit the defendants' websites. The court in Global Patent noted, however, that the patented method did not begin "until a computer user visits Defendant's website. If no person ever visited Defendant's website, then Plaintiff's patent would never be infringed." Id. The Global Patent court, therefore, did not rely on the remote user's initial choice to visit the website to establish that the client was not directed or controlled by the website owner. Instead, the court observed that there was no allegation of any facts that would render the defendant "otherwise vicariously liable for the acts of the remote user." Id. That observation is clearly correct in light of the fact that the claim in dispute contained limitations that had to be performed by a third party. That third party, who performed the volitional steps of identifying and inputting a query, was plainly acting independently of the defendant website operator. Because visiting the website was not one of the recited steps in the claim at issue in Global Patent, it was not a valid ground on which to base a finding of divided infringement. But because the volitional step of identifying and inputting a query was one of the recited steps, the independent performance of at least that step supported the district court's ruling that the infringing conduct was divided between two entities, without the requisite direction or control, and therefore could not support a judgment of infringement. The holding of that case therefore does not support the defendants' position here.

In this case, visiting the defendants' websites is not one of the recited steps of the claims of the '730 patent. Nor, unlike in Deep9 Corp. v. Barnes & Noble, Inc., is there any step in the '730 patent that recites actions akin to downloading updated material from the defendants' websites, which in Deep9 is triggered by the user of the Nook device connecting to the Internet "to set in motion the patented steps required by the asserted claims." 2012 WL 4336726, at *9; see also id. at *10 ("[T]he user of the Nook device is a necessary actor in Deep9's theory of infringement, which requires the Nook device user to connect the device to the Internet to practice the `downloading' and `updating' steps of the asserted claims."). Because the court in Deep9 regarded "the user's choice to connect his or her Nook device to the Internet," id. at 13, as essential to the performance of the claimed method, the court concluded that there was no showing in that case that the defendant controlled or directed each step of the asserted claims. In this case, by contrast, the steps of the claims begin with the encryption and transmission of the data and are followed by the steps of receipt and decryption, which, according to TQP's evidence, is performed automatically, without any discretionary action by the receiving party. Under TQP's evidence, all of the steps performed by the receiver are thus performed at the direction and under the control of the defendants' servers.

Changing tacks, the defendants note that when clients access the defendants' websites they may use different browsers, which use different encryption protocols. The defendants then argue that if the transmitting server uses OpenSSL 1.0.0, it generates or supplies a new key value one byte at a time, which RC4 combines into a 32-bit value used to encrypt 32 bits of data at one time. One of the browsers that the client might use, according to the defendants, would be Firefox 2.0, which includes an implementation that "generates or supplies a new key value that is one byte in length." For that reason, the defendants argue, the recited limitation that refers to the generation of a key value at a "time dependent on said predetermined characteristic" would not be the same at the receiver as at the transmitter.

Through their respective experts' declarations, the parties dispute whether, if the RC4 algorithm is used, the clients are directed to produce a new key value for each byte of message text. That argument, however, is not really an argument directed at divided infringement; instead, it is directed at whether particular combinations of the defendants' servers and the clients' browsers satisfy all the limitations of the asserted claims. That is, the argument does not go to whether the server exercises direction or control over the operation of the client computers. It simply presents the question whether the interaction of the servers and the client computers results in infringement in particular instances. The defendants' argument therefore does not undercut TQP's contention that there is a disputed issue as to whether the defendants' servers exercise direction or control over the functioning of the computers of clients who visit the defendants' websites and adopt a cipher suite based on the RC4 algorithm. To the extent that the defendants' evidence on that point is relevant to the issue of direction or control, it does not establish that the servers lack direction or control over the client computers; it shows only that the direction and control may take different forms depending on how the client computers are programmed.

Finally, the defendants argue that the input from the defendants' servers does not result in infringement because the clients' computers, which receive the transmitted data and decrypt it, do not perform all aspects of the three recited steps that are assigned to the receiver computer. Once again, however, that argument is not about divided infringement, but rather addresses the question whether all the limitations of the claims have been met, whether by one party alone or all parties collectively. That argument, as presented in the defendants' briefs, is essentially the same as the argument made in the defendants' motion for summary judgment of noninfringement, decided separately today. It will be addressed in that context.

For the foregoing reasons, the Court cannot rule that there is divided infringement in this case as a matter of law. Summary judgment of noninfringement on that ground is therefore denied.

* * * * *

In light of the decisions set forth in this order, denying three of the defendants' motions for summary judgment, and the decision set forth in a separate order granting the defendants' motion for summary judgment of non-infringement, the stay of proceedings previously ordered in this case is hereby vacated. The parties are directed to advise the Court within 14 days as to how they wish to proceed with this litigation.

It is so ORDERED.