Summary
explaining that, at the summary judgment stage, courts may not rely on expert reports that fail to state the basis and reasoning behind their opinions
Summary of this case from Santiago-Montanez v. United StatesOpinion
CIV. NO.: 10-1855(SCC)
01-30-2014
UPDATECOM, INC., Plaintiff, v. FIRSTBANK P.R., INC., Defendant.
OPINION AND ORDER
We have reviewed the motions for summary judgment filed in this case, Docket Nos. 374, 376, and, as we explain below, conclude that they must be denied in large part. As such, this case will proceed to trial on February 3, 2014. Nonetheless, the motions allow us to circumscribe the scope of that trial in certain regards. We explain below.
Assuming the parties' counsels familiarity with the law, we omit a discussion of the basic summary judgment standards.
A. Background
At the heart of this case is Plaintiff UpdateCom's allegation that a piece of software created by Defendant FirstBank called ATM Transaction Broker ("AAB") infringed on the copyright for a piece of software owned by UpdateCom called End2End. We note, first of all, that there is an issue of material fact as to the actual ownership of End2End. According to UpdateCom, End2End was created exclusively by UpdateCom employees and was licensed to FirstBank; according to FirstBank, however, UpdateCom only acted on a consultative basis, and the software that became End2End was created for FirstBank, which jointly or solely authored it. As this matter is contested, it can only be decided by a trial after weighing the parties' evidence.
A second matter of dispute is the date on which AAB went into effect. According to FirstBank, End2End was removed from use, and version 1.0 of AAB was installed, on September 15, 2010. UpdateCom's expert, however, states that according to his investigation, version 1.0 of AAB was not compiled until October 10, 2012, and therefore it could not have been in use before that date. See Docket No. 379, at 28. Thus, UpdateCom implies, from September 15, 2010, until at least October 10, 2012, either End2End or an End2End-like software predating the AAB version 1.0 was being used. Given this dispute, we cannot, without a trial, conclude when AAB version 1.0 was put into use.
This is in part the basis for UpdateCom's related spoliation motion.
But UpdateCom did inspect AAB version 1.0, as well as its subsequent versions, and its expert produced a report regarding whether that copied End2End. After reviewing that report, we are convinced that no reasonable factfinder could conclude that AAB versions 1.0 and later infringe upon End2End's copyright, if that copyright is indeed owned by UpdateCom.
B. The Law of Software Copyright
There are two essential elements to a claim of copyright infringement. First, the plaintiff must show "ownership of a valid copyright"; second, he must show "copying of constituent elements of the work that are original." Feist Publ'ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 361 (1991). The copying inquiry has, in turn, two components. First, the plaintiff must show that his work was "actually copied," either by "direct evidence of copying or by indirect evidence, including access to the copyrighted work, similarities that are probative of copying between the works, and expert testimony." Laureyssens v. Idea Grp., Inc., 964 F.2d 131, 140 (2d Cir. 1992). Once actual copying is shown, the plaintiff must also demonstrate that the copying was actionable, "by showing that the second work bears 'substantial similarity' to protected expression in the earlier work." Castle Rock Entm't v. Carol Publ'g Grp., 150 F.3d 132, 137 (2d Cir. 1998).
As we've said, the first of the ownership element is in dispute, and so we will focus here only on copying.
In the context of software cases, "access is either conceded or easily proved," and so the most important question is whether the two works are "substantially similar." 4 NLMMER ON COPYRIGHT § 13.03[F]. This is because even where copying is admitted, "no legal consequences will follow from that fact unless the copying is substantial." Newton v. Diamond, 388 F.3d 1189, 1193 (9th Cir. 2004). In analyzing whether two software works are substantially similar, we must also distinguish between the literal and nonliteral elements of those works. The term "literal elements" refers to a program's "source and object codes," that is, the text, in various programming or machine languages, in which the program is written by programmers or read by a computer. See Computer Assocs. Int'l, Inc. v. Altai, Inc., 982 F.2d 693, 702 (2d Cir. 1991). The term "nonliteral elements," by contrast, refers to the program's architecture and organization, including, for example, "general flow charts as well as the more specific organization of inter-modular relationships, parameter lists, and macros." Id.
We must also distinguish between literal and nonliteral copying. Literal copying is the exact copying of elements from an original work to the derivative work, and in analyzing this sort of copying we are guided by the First Circuit's opinion in Lotus Dev. Corp. v. Borland Int'l, Inc., 49 F.3d 807 (1995). Nonliteral copying, by contrast, "is copying that is paraphrased or loosely paraphrased rather than word for word." Id. at 814. We analyze nonliteral according to Altai's Abstraction-Filtration-Comparison analysis.
Regardless of the type of element or the type of copying alleged, however, we are guided by a few basic principles. Chief among these is that "copyright does not protect an idea, but only the expression of the idea." Altai, 982 F.2d at 703 (citing Baker v. Selden, 101 U.S. 99 (1879)); see also 17 U.S.C. ¶ 102(b). In this vein, we must confront the fact that computer programs are "essentially utilitarian" works that "combine[] creative and technical expression." Altai, 982 F.2d at 704 (internal quotations omitted). This matters because the processes described by a utilitarian work cannot be copyrighted. Id. Moreover, "those aspects of a work, which 'must necessarily be used as incident to' the idea, system or process that the work describes, are also not copyrightable." Id. (quoting Baker, 101 U.S. at 104); see also id. at 705 ("[T]hose elements of a computer proram that are necessarily incidental to its function are . . . unprotectable."). Essentially, this is an application of the merger doctrine, which holds that "[w]hen there is . . . only one way to express an idea, the idea and its expression are inseparable." Concrete Machinery Co. v. Classic Lawn Ornaments, Inc., 843 F.2d 600, 606 (1st Cir. 1988). This is important in the software context because "[w]hile, hypothetically there might be a myriad of ways in which a programmer may effectuate certain functions within a program . . . efficiency concerns may so narrow the practical range of choice as to make only one or two forms of expression workable options." Altai, 982 F.2d at 708. Thus, "the fact that two programs contain the same efficient structure may as likely lead to an inference of independent creation as it does to one of copying." Id. Following these principles, courts have concluded that software algorithms cannot be copyrighted because they only describe uncopyrightable methods of operation. See, e.g., Torah Soft Ltd. v. Drosnin, 136 F. Supp. 2d 276, 291 (S.D.N.Y. 2001) ("An algorithm, however, is clearly a method of operation which cannot be protected.").
Software cases also apply the scenes a faire doctrine, which "denies copyright protection to elements of a work that are for all practical purposes indispensable, or at least customary, in the treatment of a given subject matter." Coquico, Inc. v. Rodriguez-Miranda, 562 F.3d 62, 68 (1st Cir. 2009). In the computer science arena, this doctrine is used to account for the fact that "in many instances, it is virtually impossible to write a program to perform particular functions in a specific computing environment without employing standard techniques." 4 NIMMER § 13.03[F][3]. These external factors include hardware and software standards, design standards, the practices of the target industry, and general programming practices. Id.; see also Altai, 982 F.2d at 709-10 (noting that "a programmer's freedom of choice is often circumscribed by extrinsic considerations"). Where features are necessitated by such considerations, they cannot be copyrighted.
C. UpdateCom's Analysis of AAB
UpdateCom's expert analyzed the source code of four versions of AAB, and he produced a report concluding that they were copied from End2End. See Docket No. 379. UpdateCom's expert alleges literal and nonliteral copying of literal and nonliteral elements. In essence, though, UpdateCom's expert makes two claims: first, that AAB copies End2End's general architecture; and second, that AAB copies specific portions of End2End's source code. We will deal with these two claims separately.
We note that UpdateCom's expert, Angel Figueroa-Cruz, who identifies himself in his report as a "senior software developer" at "Update Computer Solutions," Docket No. 379, at 3, is in fact UpdateCom's "[p]resident and stockholder," VERIFIED COMPLAINT, Docket No. 1, ¶ 11. We find this oversight in his curriculum vitae curious and potentially problematic.
1. Alleged Copying of Source Code
UpdateCom's expert's report gives four examples of short sections of code (totaling 11 or 13 lines depending on how one counts) found in AAB version 1.0 that it says are "identical" to sections of code found in End2End. See Docket No. 379, at 36-40. FirstBank focuses on this small amount of alleged copying, arguing that even if those lines were directly copied, it would be de minimis and not actionable. UpdateCom responds by claiming that its expert did not say that these were the only lines of code that were copied. Docket No. 386, at 2 ("Contrary to defendant's allegations, FirstBank illegally copied or used thousands of lines from End2End's source code, not just 11 lines."). The problem for UpdateCom, however, is that its expert report gives no support whatsoever to the claim that AAB copied more than, at most, the short sections of code listed at pages 36 through 40 of its expert report. These supposed thousands of lines of copied code are completely unmentioned, and therefore we will not consider any claims regarding them. And even if it were established that these lines had been directly—and "identically"—copied, we would conclude that they—11 lines of 5,000, or about 0.22%—constituted de minimis copying in the absence of other evidence. See On Davis v. The Gap, Inc., 246 F.3d 152, 172 (2d Cir. 2001) ("[W]here unauthorized copying is sufficiently trivial, 'the law will not impose legal consequences.'" (quoting Ringgold v. Black Entm't Television, Inc., 126 F.3d 70, 74 (2d Cir. 1997))); cf. MiTek Holdings, Inc. v. Arce Eng'g Co., Inc., 864 F. Supp. 2d 1568, 1575, 1585 (S.D. Fla. 1994) (finding a similarity of, at most, 2% to be de minimis), aff'd, 89 F.3d 1548 (11th Cir. 1996).
Seeking application of Rule 26, FirstBank pointed out this failing. See Docket No. 391, at 2. UpdateCom responded that "[c]learly, Plaintiff Expert's Report specifically identified where to find the rest of the source code lines that were 'copied or used without' authorization." Docket No. 403-1, at 2. But it did not cite to that portion of the expert report, nor can we find it.
But UpdateCom's problems do not stop here. The few examples of "identical" source code that it has identified are not, in fact, identical; at best, they are merely similar. This is not itself a problem, as UpdateCom is entitled to allege the nonliterael copying of its source code. However, its expert report fails to make any such a showing. Figure 23, on page 37 of the expert report, and its accompanying explanation, is a representative example. In the figure, some five lines of what is said to be AAB code are compared with five lines of End2End code. There are superficial similarities between the compared lines (e.g., the first line of each code uses the phrase "byte[] RawData = new byte"), but there are also differences (e.g., after "new byte," the first line of the AAB code says "[messageLength]," while the End2End code says "[state.length]"). What the report fails to do is explain the significance of the similarities and differences. Instead, it gives a cursory explanation of what the code does (but no explanation of the specific lines) and concludes that the sections are "identical." Nothing in the report, with regard to this or the other examples, attempts to explain why the similarities are significant, and not, for example, similarities required by efficiency or external constraints. In this sense, the report determines that the AAB code infringes End2End in a purely conclusory manner. Consequently, we cannot help but find that the report's comparison of the programs' source code at pages 36 through 40 fails to meet the requirements of Federal Rule of Civil Procedure 26(a)(2)(B), which requires expert reports to state the basis and reasoning behind their opinions. Moreover, because trial is set to begin in mere days, the only available remedy is striking this portion of the expert report. See Santiago-Diaz v. Laboratorio Clinico y de Referenda del Este, 456 F.3d 272, 276 (1st Cir. 2006) (affirming district court's striking of a conclusory expert report, and holding that the "baseline" sanction for such a failure is preclusion of the evidence); Salgado v. Gen. Motors Corp., 150 F.3d 735, 742 n.6 (7th Cir. 1998) ("Expert reports must not be sketchy, vague, or preliminary in nature."); Kerlinisky v. Sandoz Inc., 783 F. Supp. 2d 236, 242 (D. Mass. 2011) (striking expert report that failed to "provide with any reasonable degree of specificity the basis and reasons for [the expert's] opinions"); Elder v. Tanner, 205 F.R.D. 190 (E.D. Tex. 2001) (striking report under Rule 26 for failing to discuss the expert's reasoning and thought process, and under Daubert for making "conclusory statements" that lacked "any elaboration or reasoning"); Fidelity Nat'l Title Ins. Co. of N.Y. v. Intercounty Nat'l Title Ins. Co., Civ. No. 00-5658, 2001 WL 789218, at *3 (N.D. Ill. July 12, 2001) ("Rule 26(a) requires that the expert report contain the basis for each opinion.").l
Necessarily, then, UpdateCom's expert will be precluded from testifying on these matters at trial.
A final problem with these examples is that, even according to the expert report and the expert's own deposition testimony, the relevant sections of code each seem to accomplish rather specific tasks (such as communicating with another vendor's software), and, as such, are likely uncopyrightable under the merger doctrine or as mere algorithms.
For all of these reasons, we find that no issue of material fact exists with regard to FirstBank's AAB version 1.0's alleged copying of literal elements from End2End.
2. Alleged Copying of Architectural Elements
UpdateCom also alleges that AAB version 1.0 copies End2End's architecture. It does this in two ways. First, the report provides a diagram of AAB's architecture. Docket No. 379, at 25. This diagram, which notably is completely unexplained in the report's text, shows, in an extremely general way, what AAB does. AAB, which is identified by a single box, seems to communicate with various other pieces of software, including Evertec's ATH network and the bank's financial software. See id. Though the report fails to actually make this comparison, the diagram is similar (though not identical) to a basic diagram of End2End's functions elsewhere in the report. See id. at 9. In both instances, there is only a single box representing the software at issue. No detail is provided as to how the software accomplishes any task; instead, the diagrams merely show that End2End and AAB allow interactions between other, disparate pieces of software. Given that the programs are not described at all, beyond their function, it hardly even seems right to refer to these diagrams as "architectural." More to the point, UpdateCom cannot possibly be claiming to have a copyright of this highly generalized architecture, which is really nothing more than an idea for a piece of software, and therefore uncopyrightable. The diagram, then, is neither evidence nor explanation for the report's claim that the AAB architecture "follows the same path as the End2End Software Architecture." Id. at 25.
Second, UpdateCom's expert relies on a table, supported by three confusing and non-contextualized images, see id. at 26-27, which it says shows that the AAB "architecture is identical to the End2End Architecture," id. at 26. The table shows eight items under the heading "Description" that it says are identical between End2End and AAB. See id. Or at least that's what we think it shows; in fact, the table goes unelaborated upon. To show why this matters, we will give one example. Under "Description," one item is "Mainframe Queue Name," and, for both End2End and AAB, it is said to be "FB.LX.QMI." But without knowing what "Mainframe Queue Name" refers to, we cannot determine whether it is an important or relevant piece of the software's architecture (or a part of the architecture at all). And, of course, the meaning of "FB.LX.QMI" is far from apparent; the report makes it impossible to determine whether such a name is, for example, required by the tasks that the software both perform. Ultimately, the report's discussion of the programs' architecture is nothing more than unexplained, unsupported conclusions. As such, we must strike the section of the report dealing with the architecture comparison, and we must conclude that no issue of fact exists as to whether version 1.0 of AAB's architecture is substantially similar to End2End's.
Moreover, because these are the only sections of the report that deal with whether AAB versions 1.0 and later are substantially similar to End2End, we must conclude that no material issue of fact exists with regard to those programs' substantial similarity. We therefore conclude, as a matter of law, that the programs are not substantially similar. As such, and without regard to whether any actual copying occurred, we conclude that AAB versions 1.0 and later do not infringe any copyright that UpdateCom has in End2End.
3. When did AAB Version 1.0 Go Into Effect?
To determine the effect of our conclusion that AAB version 1.0 does not infringe End2End, it is necessary to know when, exactly, AAB version 1.0 went into production. According to FirstBank's statement of uncontested material facts, that date was September 15, 2010. See Docket No. 375, ¶¶ 25 (version 1.0 was the first version of AAB), 19 (AAB installed on September 15, 2010). Relying on its expert's report and its most recent spoliation motion, Docket No. 378, UpdateCom disputes this date. Docket No. 387, ¶¶ 19, 25. To determine whether UpdateCom has successfully opposed FirstBank's proposed fact, we must look at the treatment of AAB's production date in UpdateCom's expert report and spoliation motion.
a. UpdateCom's Fourth Motion Regarding Spoliation
For some time now, UpdateCom has been arguing that FirstBank has willfully destroyed the evidence that would prove its liability to UpdateCom. In June 2013, we denied UpdateCom's first spoliation motion, which we found to lack support, but ordered FirstBank to produce to UpdateCom all versions of the AAB source code. Docket No. 276. UpdateCom then sought reconsideration of our Order. Docket No. 279. We denied reconsideration because we found that it requested relief that should have been requested in the first motion, but which was not. Docket No. 282. UpdateCom then filed a purported "second" motion regarding spoliation, Docket No. 290; it, too, was denied, because it repeated the arguments of the first motion and the reconsideration of that motion's denial, and it was, in that sense, a motion for reconsideration of our denial of reconsideration. Docket No. 306. UpdateCom raised spoliation issues again during a status conference held on August 7, 2013; at that time, the Court made clear that it would only consider arguments based on evidence that was not available when the previous several motions were filed. Docket No. 325, at 5.
With that background in mind, we take up UpdateCom's most recent spoliation motion. Docket No. 378. As stated in the motion, UpdateCom's belief is that on September 15, 2010, FirstBank did not implement AAB; instead, it implemented a slightly modified version of End2End that incorporated certain changes required by federal regulations. See id. at 3. UpdateCom contends that this modified software infringed on its copyright. See id. What prompted its motion, though, was newly-acquired evidence suggesting that early versions of AAB either still exist or were destroyed very late in this litigation.
The basis for this belief is UpdateCom's deposition, on September 4, 2013, of Omar Cruz-Salgado, formerly the project manager in charge of developing AAB for FirstBank. See Docket No. 378-1. After testifying that AAB began production in April 2010 and first went into production on September 15 of that year, id. at 33, Cruz testified that he would save pre-production drafts of the software onto his work laptop as well as on FirstBank's version control system, id. at 34. Furthermore, he testified that those drafts might still exist on FirstBank's computers—and that he believed that they existed at least through when he left the Bank, in October 2012, two years after this litigation began. Id. at 43-44. After these revelations, FirstBank's counsel promised to look into whether these pre-productions drafts still existed, but none, apparently, were ever turned over.
To be sure, these facts call into question whether FirstBank has diligently and fully complied with its discovery obligations in this case. Their potential for impact on this portion of the case's substance, however, is much smaller. This is because the new evidence relates solely to the existence pre-production copies of AAB. Cruz's deposition testimony does not suggest that any of these pre-production versions were ever implemented, and, to the contrary, it fully supports FirstBank's position that AAB version 1.0 was implemented on September 15, 2010. And given our conclusion that the production version of AAB did not actionably infringe any copyright that might be held by UpdateCom, we do not see how this new evidence can support the remedies that UpdateCom seeks. To the contrary, even if these pre-production versions—essentially, drafts—contained substantial, literal copying of End2End's source code, that would only be evidence of actual copying. UpdateCom's claim would still fail the substantial similarity prong for the reasons explained above, and, accordingly, so would its infringement claim. Therefore, we find that no spoliation remedies are warranted with respect to this claim, and we deny UpdateCom's motion in this regard.
We take up this matter in more detail below.
See, e.g., Walker v. Time Life Films, Inc., 784 F.2d 44, 52 (2d Cir. 1986) ("[E]arly drafts might be useful to show that defendants had gained access to plaintiff's work, borrowed from it, and later made changes in order to conceal that borrowing. . . . However, since we conclude as a matter of law that . . . no substantial similarity exists between the protectible portions of the final versions the works, any error in the exclusion of the early drafts was harmless."); Quirk v. Sony Pictures Entm't Inc., Civ. No. 11-3773, 2013 WL 1345075 (N.D. Cal. Apr. 2, 2013) (finding early drafts irrelevant where only issue was substantial similarity); Flaherty v. Filardi, Civ. No. 03-2167, 2009 WL 749570, 11* (S.D.N.Y. Mar. 20, 2009) ("[T]he contents of screenplay drafts that are not reflected in the finished motion picture are not relevant to the substantial similarity analysis . . . ."), aff'd, 460 F. App'x 66 (2d Cir. 2012); Flaherty v. Filardi, Civ. No. 03-2167, 2007 WL 2734633, *4 (S.D.N.Y. Sept. 19, 2011) (holding that because the defendant's ultimate work was not substantially similar to the plaintiff's, "the drafts prepared in the course of . . . development and production constitute interim drafts of a published non-infringing work, and are not actionable under the Copyright Act").
As another piece of evidence in support of its motion, UpdateCom alleges that another FirstBank employee, Eric Lopez, testified during his deposition that he was not sure whether AAB was placed into production on September 15, 2010. Docket No. 378, at 19. However, in the very portion of the deposition that UpdateCom cites for this proposition, Lopez testified that he personally installed AAB on that date. Docket No. 378-3, at 30. As a last piece of evidence, the motion cites the portion of expert report that we discuss in the next section.
A large portion of the motion concerns purported new evidence that FirstBank did not inform its employees of their duty to preserve evidence. Again, this could be the proper subject of sanctions, but given the evidence that is missing, we do not think it would warrant the remedies that UpdateCom requests.
b. The Expert Report
Apart from an inference based on spoliation, the only basis for UpdateCom's denial that AAB was implemented on September 15, 2010, is found in its expert report. Specifically, UpdateCom refers to a two-page passage in the report that discusses a "Software Design Document" dated September 20, 2010. See Docket No. 379, at 44-45. According to UpdateCom, this passage supports the proposition that "[o]n September 15, 2010, FirstBank started the Design of [AAB] and was finished on September 20, 2010." Docket No. 387, ¶ 19 (emphasis added). This is, in fact, quite contrary to what the report actually states. In fact, the report says simply that software design documents generally provide "guidance for the development team," but that in this case, the document "was not used as guidance" because "it was created after" September 15, 2010. Docket No. 379, at 44. The report does not purport to conclude, based on this document, that the development of AAB began on September 15, 2010. Indeed, elsewhere the report concludes that work on AAB began at least as early as March 2010. See id. at 31 ("The FirstBank development of [AAB] source code started at least on March 25 2010 . . . .").
UpdateCom's also claims that Cruz's deposition testimony supports the fact that development of AAB began on September 15, 2010. See Docket No. 387, ¶ 19. In reality, Cruz testified to exactly the opposite point: that the design document was created after AAB was implemented. Docket No. 378-1, at 86.
For this reason alone, we should deem admitted FirstBank's proposed fact that AAB was implemented on September 15, 2010. However, for the sake of completeness we will deal briefly with the expert report's other suggestion that AAB was not implemented on September 15, 2010. At two different points, the report suggests that AAB version 1.0 could not have been implemented on September 15, 2010, because it was not compiled until October 10, 2012. See Docket No. 379, at 28, 35. Compiling is the process of turning source code, which is written by a human, to another language, like object code, that can be read by a computer. To support his contention that AAB version 1.0 was not compiled until October 10, 2012, UpdateCom's expert relies completely on two figures. Each figure seems to be an image of a Windows file manager. On the left side of each figure is an extensive hierarchy of file folders; on the right are the contents of a specific folder, showing other files and file folders. In each case, certain specific files are surrounded by a drawn box, and these files have, according to the file manager, a modified and created date of October 10, 2012, or later. See id. at 29-30, 35. The problem, though, is that the report fails to explain why these file folders and these files are significant in any respect. Moreover, it fails to explain why the creation date, as registered by the Windows file manager, matters—or even whether it is an appropriate basis on which to determine the date that AAB was compiled. In short, this section of the report suffers from precisely the same failing that do the sections we've previously discussed: it reaches conclusions without giving any real description of the expert's thought process or reasoning. And as with the other sections, not only can these conclusions not create an issue of fact as to the date that AAB was implemented, the entire section must be struck as in violation of Rule 26.
4. Wrapping Up: Does AAB Infringe End2End?
For all of these reasons we conclude that AAB version 1.0 was implemented on September 15, 2010, and that it does not infringe any copyright that UpdateCom has in End2End. For all of the same reasons, we deem unopposed the fact that up until September 15, 2010, FirstBank was running an unmodified version of End2End. See Docket No. 375, ¶ 17 (proposed fact); Docket No. 387, ¶ 17 (repeating the substance of the objection to fact number 19).
D. Other Infringement Claims
UpdateCom alleges that it is the sole author of End2End, in which it granted a non-exclusive license to FirstBank. We find that there is a genuine issue of material fact as to authorship and, therefore, also as to the existence of such a license. For this reason, summary judgment is not possible regarding UpdateCom's claims that FirstBank infringed its copyright by using End2End inconsistently with the license agreement. We note, however, that contrary to FirstBank's arguments that such claims sound only in contract, we find such use, if proved at trial, might constitute infringement. We are not convinced, though, that all violations of a licensing agreement constitute actionable infringement. For example, while use of the product after the cancellation of the license, and perhaps certain copying in violation of the agreement is actionable, reverse engineering may not be. See, e.g., Davidson & Assocs., Inc. v. Internet Gateway, Inc., 334 F. Supp. 2d 1164, 1180 (E.D. Mo. 2004) ("Reverse engineering as fair use is firmly established.").
Of course, "private parties are free to contractually forego the limited ability to reverse engineer a software product under the exemptions of the Copyright Act." Bowers v. Baystate Techs., Inc., 320 F.3d 1317, 1325-26 (Fed. Cir. 2003) (applying First Circuit law). But this does not mean that the reverse engineering would be actionable infringement rather than breach of contract.
We also think that a fact issue remains as to UpdateCom's infringement allegations related to the IVR software. As we understand UpdateCom's allegations, FirstBank used IVR software until at least October 2011, when the bank replaced it with a new piece of software developed in-house. UpdateCom does not seem to claim that FirstBank infringed a copyright on IVR; instead, its contention seems to be that if IVR ran until October 2011, so must have at least some portion of End2End, because End2End provided—and AAB did not—the backend that allowed IVR to communicate with other software, includ- ing Evertec's and FirstBank's mainframes. See Docket No. 393, at 8-9. Issues of fact exist as to whether IVR requires End2End to function, as well as when IVR was removed from use.
That said, we have serious doubts about whether UpdateCom's expert's report's opinions regarding the necessity of End2End for IVR to function are anything more than conclusory. Indeed, the passages that UpdateCom cites only state bald conclusions, without any indication of how they were reached. See, e.g., Docket No. 376, at 18 (citing Docket No. 379, at 4, 6). We find it unnecessary to decide this matter here in this opinion, but we nonetheless reserve the right to strike this portion of the report and preclude UpdateCom's expert's testimony on this subject.
II. Other Claims
UpdateCom also makes claims under two state-law causes of action. First, it claims that FirstBank violated Puerto Rico's moral rights statute, P.R. Laws Ann. tit. 31, § 1401. Second, UpdateCom makes a claim for collection of monies allegedly owed to it by First Bank. See P.R. Laws Ann. tit. 31, ¶ 3373. Both of these claims are related to the matters of authorship and license, and, as such, summary judgment would be inappropriate. As such, these claims will go to trial.
We note, moreover, that we believe that some of the allegedly spoliated evidence—especially the previous drafts of AAB—may have some relevance to all of these claims insofar as it might have shed light on whether FirstBank had reverse-engineered End2End or had otherwise copied it, thus violating the license agreement. We are therefore considering giving the jury a limited adverse inference instruction. Before we decide whether such an instruction is warranted, however, we require that FirstBank directly answer certain questions: Do pre-production drafts of AAB still exist? If so, have they been turned over? (And if not, why not?) If they do not, when were they deleted, both from the version control system and the programmers' laptops? And how did this occur after legal hold letters were sent? In addition, FirstBank should provide any other relevant factual information about its destruction or withholding of this and related evidence. An informative motion on this subject shall be filed before the start of trial.
We read certain language in FirstBank's filings as suggesting that it believed it was not required to produce anything but production versions of AAB. See Docket No. 392, at 4-5. We are inclined to think that this would constitute a willful misreading of our previous discovery orders.
III. Other Factual Matters
In order to streamline the upcoming trial, and based on the parties' factual submissions at the summary judgment stage, we make the following factual findings which, pursuant to Rule 56(g), shall govern this case.
It is undisputed that on December 10, 2007, version 1.0 of End2End went live in FirstBank's systems. Docket No. 375, at ¶ 6; Docket No. 387, at ¶ 6. Subsequently, other versions were produced between July 2008 and February 2010. Docket No. 375, ¶ 12; Docket No. 387, ¶ 12. The last of these, called "Version 2.8 (TISA/TILA)," was installed by UpdateCom in February 2010. Docket No. 375, ¶ 13; Docket No. 387, ¶ 13. Then, on April 21, 2010, UpdateCom delivered to FirstBank the source code for that version of End2End. Docket No. 375, ¶ 14; Docket No. 387, ¶ 14. FirstBank's personnel were unable to compile this version of End2End. Docket No. 375, ¶ 15. Subsequently, Angel Figueroa of UpdateCom send FirstBank a guide to compiling the source code for Version 2.8 (TISA/TILA). Docket No. 375, ¶ 16.
UpdateCom purports to deny this fact, but fails to actually controvert the fact's substance. Instead, UpdateCom's putative denial simply adds other, additional facts. See Docket No. 387, ¶ 13. We therefore deem the fact admitted.
UpdateCom purports to deny this fact, but its denial is merely an allegation that the attempt to compile End2End violated the purported license. See Docket No. 387, ¶ 15. As such, it fails to controvert the proposed fact's substance, and the proposed fact is deemed admitted.
UpdateCom purports to deny this fact, but its denial is merely an argument that the guide was sent for reasons other than helping FirstBank compile Version 2.8 (TISA/TILA). See Docket No. 387, ¶ 16. That argument, moreover, is not supported by the exhibits on which UpdateCom relies. Moreover, in its own statement of material fact in support of its motion for summary judgment, UpdateCom proposes substantially the same fact. Docket No. 377, ¶ 21 ("Figueroa of Updatecom provided instructions to FirstBank on how to compile the source code of version 2.8 of End2End and asserted that there should not be a problem with compiling the End2End source code if FirstBank followed those instructions."). The fact is deemed admitted.
In its statement of uncontested material facts in support of its own motion for summary judgment, UpdateCom proposes that it "fulfilled its part of the Agreement with FirstBank and performed all tasks that were required by FirstBank and 3 of the pending 4 invoices, are overdue for more than four (4) years. One invoice is overdue around 3 years." Docket No. 377, ¶ 29. Furthermore, it claims that FirstBank owes it $108,750, plus interest. Id. ¶ 30. FirstBank purports to deny these facts, but it does so only through quotations from its counterclaim and answer to the complaint. See Docket No. 383, ¶¶ 29-30. But "pleadings are not evidence properly considered on motions for summary judgment." In re Rezulin Liability Litig., Civ. No. 00-2843(LAK), 2005 WL 713331, at *1 (S.D.N.Y. Mar. 29, 2005); see also Heggem v. Kenney, Civ. No. 08-437(RAJ), 2009 WL 2486136, at *2 (W.D. Wash. Aug. 12, 2009) ("Conclusory allegations in pleadings are not evidence, and cannot by themselves create a genuine issue of material fact." (citing Project Release v. Prevost, 722 F.2d 960, 969 (2d Cir. 1983))). We therefore deem these facts admitted.
We do not believe that these admitted facts are, themselves, sufficient to permit the court to enter summary judgment on any claims. Moreover, we are not concluding that FirstBank "owes" any amount in a legal sense.
--------
IV. Conclusion
For all of the reasons stated above, UpdateCom's motion for summary judgment, Docket No. 376, is DENIED, and FirstBank's motion, Docket No. 374, is GRANTED IN PART and DENIED IN PART consistent with this opinion.
As we see it, none of UpdateCom's claims are dismissed in their entirety. Nonetheless, the parties are forbidden from offering any evidence or argument at trial inconsistent with this Order's findings. See Fed. R. Civ. P. 56(g) (permitting a court to "enter an order stating any material fact . . . that is not genuinely in dispute and treating the fact as established in the case"). Specifically, we conclude:
1. Pursuant to Rules 26 and 37, the following portions of UpdateCom's expert report are stricken: pages 25 through 27, regarding AAB's architecture; pages 28 through 30 and page 35, regarding AAB's creation date; and pages 36 through 40, comparing AAB's and End2End's source codes. UpdateCom's expert is precluded from testifying about any of these matters. Moreover, because of the substantial failings we have identified in the expert report, UpdateCom must SHOW CAUSE why its expert's report should not be stricken, and its expert's testimony precluded, in its entirety.
2. End2End version 1.0 was installed on FirstBank's systems on December 10, 2007. Between July 2008 and February 2010, subsequent versions were also installed.
3. In February 2010, UpdateCom installed on FirstBank's systems End2End version 2.8 (TISA/TILA). The source code for this version was delivered to FirstBank by UpdateCom on April 21, 2010.
4. FirstBank's personnel could not compile End2End version 2.8. Subsequently, Angel Figueroa of UpdateCom provided FirstBank with a guide to compiling that version of End2End.
5. FirstBank installed version 1.0 of AAB on September 15, 2010.
6. Before September 15, 2010, FirstBank was running an un-modified version of End2End.
7. AAB versions 1.0 and later do not copy any literal elements from End2End.
8. AAB versions 1.0 and later do not copy End2End's architecture in an actionable way.
9. AAB versions 1.0 and later are not substantially similar to End2End.
10. AAB versions 1.0 and later do not infringe any copyright FirstBank might hold in End2End.
11. UpdateCom performed all services required by its agreement with FirstBank, and it has four pending, unpaid invoices. Three of these have been overdue for more than four years; one has been overdue for more than three years. The total amount of money owed on these invoices is $108,750.
IT IS SO ORDERED.
In San Juan, Puerto Rico, this 30th day of January, 2014.
SILVIA CARREÑO-COLL
UNITED STATES MAGISTRATE JUDGE