IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF DELAWARE THOMSON REUTERS ENTERPRISE CENTRE GMBH and WEST PUBLISH- ING CORP., Plaintiffs, v. No. 1:20-cv-613-SB ROSS INTELLIGENCE INC., Defendant.
Jack B. Blumenfeld, Michael J. Flynn, MORRIS, NICHOLS, ARSHT & TUNNELL LLP, Wil- mington, Delaware; Dale M. Cendali, Jeremy King, Eric A. Loverro, Joshua L. Simmons, KIRKLAND & ELLIS LLP, New York, New York; Miranda Means, KIRKLAND & ELLIS LLP, Boston, Massachusetts; Yungmoon Chang, Allyn Belusko, KIRKLAND & ELLIS LLP, Los Angeles, California. Counsel for Plaintiffs. David Ellis Moore, Bindu Ann George Palapura, Andrew L. Brown, POTTER ANDERSON & CORROON LLP, Wilmington, Delaware; Warrington Parker, Joachim B. Steinberg, Ja- cob Canter, Christopher J. Banks, Anna Z. Saber, CROWELL & MORING LLP, San Fran- cisco, California; Mark A. Klapow, Crinesha B. Berry, Lisa Kimmel, CROWELL & MOR- ING LLP, Washington, D.C.; Emily T. Kuwahara, Jordan Ludwig, CROWELL & MORING LLP, Los Angeles, California; Ryan Henry Seewald, CROWELL & MORING LLP, Denver, Colorado. Counsel for Defendant. MEMORANDUM OPINION August 02, 2024 BIBAS, Circuit Judge, sitting by designation.
Not all expert testimony is admissible at trial. Each party has moved to exclude portions from two of the other side’s experts’ opinions relevant to this case’s copyright issues. I previously reserved ruling on these motions in whole or in part. D.I. 549, 550. Now that trial approaches, I consider them under Federal Rule of Evidence 702 and the Daubert standard. See Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993). Qualified experts’ opinions are admissible if they are based on reliable meth- odology and fit the facts of the case. Elcock v. Kmart Corp., 233 F.3d 734, 741 (3d Cir. 2000). At this point, the experts’ qualifications are not at issue.
I. Jonathan Krein used a reliable methodology Ross challenges Jonathan Krein’s methodology for figuring out whether headnotes owned by Thomson Reuters are substantially similar to data used by Ross to train its AI. D.I. 268, Krein Opening Br. 9–12. It says that Krein’s methodology is unreliable both because of what he included in the dataset and because of how he evaluated linguistic similarity within the dataset. Id. But Krein’s methodology is reliable. In his reply expert report, he details how he
selected the data that he used to decide substantial similarity. D.I. 266-1, at 735–39. Though his analysis may not be perfectly replicable because it involves some qualita- tive judgments, that does not make it improper. Courts require only reliability, not that experts use the “best methodology or unassailable research.” In re TMI Litig., 193 F.3d 613, 665 (3d. Cir. 1999). And there are no obvious flaws that would render his methodology unreliable or confuse a jury. Because Krein explained his approach and it is reasonable, I deny Ross’s motion to exclude. II. James Malackowski’s opinions are helpful and not impermissible legal analysis Ross also seeks to exclude James Malackowski’s opinions related to the statutory damages available to Plaintiffs if they prevail on the copyright issues. D.I. 349, Malackowski Reply Br., at 2. It argues that his opinion reflects either mathematics
too simple to require an expert or impermissible legal analysis. Id. Not so. Copyright damages can be awarded either as actual damages plus in- fringer profits or as statutory damages. 17 U.S.C. § 504. Malackowski devotes much of his report to calculating actual damages. D.I. 275-1, at 48–60. In one small portion, he also selects facts from other expert reports to calculate statutory damages under multiple scenarios. D.I. 275-1, at 62–63. “[S]imple math” alone is insufficient. All-
scripts Healthcare, LLC v. Andor Health, LLC, No. 21-704, 2022 WL 3021560, at *5 (D. Del. July 29, 2022). But Malackowski calculates statutory damages because stat- utory damages are the alternative to actual damages. Plus, because he writes condi- tionally, he does not impermissibly “testify as to the governing law of the case” or “render[ ] a legal opinion.” Berckeley Inv. Grp., Ltd. v. Colkitt, 455 F.3d 195, 217 (3d Cir. 2006). So I deny Ross’s other motion to exclude. III. Admitting Barbara Frederiksen-Cross’s opinions does not prejudice Plaintiffs Plaintiffs argue that Barbara Frederiksen-Cross introduced new arguments in her reply expert report. D.I. 261, Frederiksen-Cross Opening Br., at 1. If she did, that would make the report untimely. See Withrow v. Spears, 967 F. Supp. 2d 982, 1002– 03 (D. Del. 2013). But untimely reports need not always be excluded. To decide that question, I consider five factors: (1) surprise or prejudice to the moving party; (2) abil- ity of the moving party to cure prejudice; (3) disruption to trial; (4) bad faith by the
nonmoving party; and (5) the importance of the evidence. LabMD Inc. v. Boback, 47 F.4th 164, 189 (3d Cir. 2022) (citing Meyers v. Pennypack Woods Home Ownership Ass’n, 559 F.2d 894 (3d Cir. 1977)). Here, any potential untimeliness is harmless. Any surprise or prejudice should have been minor because Frederiksen-Cross’s report was filed almost two years be- fore trial. See D.I. 223. And surprise or prejudice may be cured by deposition. See In re Paoli R.R. Yard PCB Litig., 35 F.3d 717, 792–93 (3d Cir. 1994). But Plaintiffs made
no effort to address any surprise when they deposed her after getting her reply report. What is more, there is no showing of bad faith here by Ross. Plus, “the most signifi- cant factor”—importance of the evidence—favors admitting her opinions because they go to the copyright issues at heart of this case. LabMd Inc., 47 F.4th at 189 (citing ZF Meritor, LLC v. Eaton Corp., 696 F.3d 254, 298 (3d Cir. 2012)). I deny Plaintiffs’ mo- tion.
IV. Alan Cox’s opinions on disgorgement and lost profits are unreliable and thus excluded Plaintiffs next seek to exclude as hypothetical and unreliable Alan Cox’s opinions on (1) disgorgement due to unjust enrichment and (2) calculation of lost profits. The owner of an infringed copyright may “recover the actual damages suffered … as a result of the infringement, and any profits of the infringer that are attributable to the infringement.” 17 U.S.C. § 504(b). To calculate the infringer’s profits, Plaintiffs need prove only Ross’s “gross revenue.” Id. Ross must then “prove … deductible expenses and the elements of profit attributable to factors other than the copyrighted work,” so they can be subtracted from gross revenue. Id.
Disgorgement. Cox’s opinion subtracts more than deductible expenses and other actual elements of profit because he also subtracts “revenues ROSS would have earned absent (‘but-for’) the alleged conduct.” D.I. 266-1, at 283. As Cox explains, “[c]ounterfactual (or but-for) profits … must take into account alternative actions that the defendant might take that did not involve (in this case) copyright infringement.” Id. at 286.
Free access — add to your briefcase to read the full text and ask questions with AI
IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF DELAWARE THOMSON REUTERS ENTERPRISE CENTRE GMBH and WEST PUBLISH- ING CORP., Plaintiffs, v. No. 1:20-cv-613-SB ROSS INTELLIGENCE INC., Defendant.
Jack B. Blumenfeld, Michael J. Flynn, MORRIS, NICHOLS, ARSHT & TUNNELL LLP, Wil- mington, Delaware; Dale M. Cendali, Jeremy King, Eric A. Loverro, Joshua L. Simmons, KIRKLAND & ELLIS LLP, New York, New York; Miranda Means, KIRKLAND & ELLIS LLP, Boston, Massachusetts; Yungmoon Chang, Allyn Belusko, KIRKLAND & ELLIS LLP, Los Angeles, California. Counsel for Plaintiffs. David Ellis Moore, Bindu Ann George Palapura, Andrew L. Brown, POTTER ANDERSON & CORROON LLP, Wilmington, Delaware; Warrington Parker, Joachim B. Steinberg, Ja- cob Canter, Christopher J. Banks, Anna Z. Saber, CROWELL & MORING LLP, San Fran- cisco, California; Mark A. Klapow, Crinesha B. Berry, Lisa Kimmel, CROWELL & MOR- ING LLP, Washington, D.C.; Emily T. Kuwahara, Jordan Ludwig, CROWELL & MORING LLP, Los Angeles, California; Ryan Henry Seewald, CROWELL & MORING LLP, Denver, Colorado. Counsel for Defendant. MEMORANDUM OPINION August 02, 2024 BIBAS, Circuit Judge, sitting by designation.
Not all expert testimony is admissible at trial. Each party has moved to exclude portions from two of the other side’s experts’ opinions relevant to this case’s copyright issues. I previously reserved ruling on these motions in whole or in part. D.I. 549, 550. Now that trial approaches, I consider them under Federal Rule of Evidence 702 and the Daubert standard. See Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993). Qualified experts’ opinions are admissible if they are based on reliable meth- odology and fit the facts of the case. Elcock v. Kmart Corp., 233 F.3d 734, 741 (3d Cir. 2000). At this point, the experts’ qualifications are not at issue.
I. Jonathan Krein used a reliable methodology Ross challenges Jonathan Krein’s methodology for figuring out whether headnotes owned by Thomson Reuters are substantially similar to data used by Ross to train its AI. D.I. 268, Krein Opening Br. 9–12. It says that Krein’s methodology is unreliable both because of what he included in the dataset and because of how he evaluated linguistic similarity within the dataset. Id. But Krein’s methodology is reliable. In his reply expert report, he details how he
selected the data that he used to decide substantial similarity. D.I. 266-1, at 735–39. Though his analysis may not be perfectly replicable because it involves some qualita- tive judgments, that does not make it improper. Courts require only reliability, not that experts use the “best methodology or unassailable research.” In re TMI Litig., 193 F.3d 613, 665 (3d. Cir. 1999). And there are no obvious flaws that would render his methodology unreliable or confuse a jury. Because Krein explained his approach and it is reasonable, I deny Ross’s motion to exclude. II. James Malackowski’s opinions are helpful and not impermissible legal analysis Ross also seeks to exclude James Malackowski’s opinions related to the statutory damages available to Plaintiffs if they prevail on the copyright issues. D.I. 349, Malackowski Reply Br., at 2. It argues that his opinion reflects either mathematics
too simple to require an expert or impermissible legal analysis. Id. Not so. Copyright damages can be awarded either as actual damages plus in- fringer profits or as statutory damages. 17 U.S.C. § 504. Malackowski devotes much of his report to calculating actual damages. D.I. 275-1, at 48–60. In one small portion, he also selects facts from other expert reports to calculate statutory damages under multiple scenarios. D.I. 275-1, at 62–63. “[S]imple math” alone is insufficient. All-
scripts Healthcare, LLC v. Andor Health, LLC, No. 21-704, 2022 WL 3021560, at *5 (D. Del. July 29, 2022). But Malackowski calculates statutory damages because stat- utory damages are the alternative to actual damages. Plus, because he writes condi- tionally, he does not impermissibly “testify as to the governing law of the case” or “render[ ] a legal opinion.” Berckeley Inv. Grp., Ltd. v. Colkitt, 455 F.3d 195, 217 (3d Cir. 2006). So I deny Ross’s other motion to exclude. III. Admitting Barbara Frederiksen-Cross’s opinions does not prejudice Plaintiffs Plaintiffs argue that Barbara Frederiksen-Cross introduced new arguments in her reply expert report. D.I. 261, Frederiksen-Cross Opening Br., at 1. If she did, that would make the report untimely. See Withrow v. Spears, 967 F. Supp. 2d 982, 1002– 03 (D. Del. 2013). But untimely reports need not always be excluded. To decide that question, I consider five factors: (1) surprise or prejudice to the moving party; (2) abil- ity of the moving party to cure prejudice; (3) disruption to trial; (4) bad faith by the
nonmoving party; and (5) the importance of the evidence. LabMD Inc. v. Boback, 47 F.4th 164, 189 (3d Cir. 2022) (citing Meyers v. Pennypack Woods Home Ownership Ass’n, 559 F.2d 894 (3d Cir. 1977)). Here, any potential untimeliness is harmless. Any surprise or prejudice should have been minor because Frederiksen-Cross’s report was filed almost two years be- fore trial. See D.I. 223. And surprise or prejudice may be cured by deposition. See In re Paoli R.R. Yard PCB Litig., 35 F.3d 717, 792–93 (3d Cir. 1994). But Plaintiffs made
no effort to address any surprise when they deposed her after getting her reply report. What is more, there is no showing of bad faith here by Ross. Plus, “the most signifi- cant factor”—importance of the evidence—favors admitting her opinions because they go to the copyright issues at heart of this case. LabMd Inc., 47 F.4th at 189 (citing ZF Meritor, LLC v. Eaton Corp., 696 F.3d 254, 298 (3d Cir. 2012)). I deny Plaintiffs’ mo- tion.
IV. Alan Cox’s opinions on disgorgement and lost profits are unreliable and thus excluded Plaintiffs next seek to exclude as hypothetical and unreliable Alan Cox’s opinions on (1) disgorgement due to unjust enrichment and (2) calculation of lost profits. The owner of an infringed copyright may “recover the actual damages suffered … as a result of the infringement, and any profits of the infringer that are attributable to the infringement.” 17 U.S.C. § 504(b). To calculate the infringer’s profits, Plaintiffs need prove only Ross’s “gross revenue.” Id. Ross must then “prove … deductible expenses and the elements of profit attributable to factors other than the copyrighted work,” so they can be subtracted from gross revenue. Id.
Disgorgement. Cox’s opinion subtracts more than deductible expenses and other actual elements of profit because he also subtracts “revenues ROSS would have earned absent (‘but-for’) the alleged conduct.” D.I. 266-1, at 283. As Cox explains, “[c]ounterfactual (or but-for) profits … must take into account alternative actions that the defendant might take that did not involve (in this case) copyright infringement.” Id. at 286. So Cox focused on hypothetical rather than actual profits. Yet that is not the standard. The statute requires subtracting actual expenses and
actual profits due to non-infringement. § 504(b). There is nothing in the text that lets parties construct and subtract hypothetical profits they might have earned had they never infringed. See Oracle Am., Inc. v. Google Inc., No. 10-03561, 2016 WL 1743154, at *3–4 (N.D. Cal. May 2, 2016). So I grant the motion to exclude this portion of Cox’s opinion. Lost profits. Lost profits are relevant to actual damages under § 504(b). If the jury
finds that Ross infringed Thomson Reuters’s copyright, actual damages include the lost profits from a hypothetical license to use the data that Ross used. Cox equates the data at issue in this case (legal questions and answers) with the data in English speech datasets, which are used to train AI models on the English language. See D.I. 266-1, at 293. But that is a different, less complicated matter than training an AI model to answer complex legal questions. The former requires only language; the lat- ter requires language and legal information. Cox does not sufficiently acknowledge, explain, or account for the differences be-
tween the datasets. Yet Ross has the burden to “affirmatively show” that the dataset licenses “are actually comparable.” M2M Sols. LLC v. Enfora, Inc., 167 F. Supp. 3d 665, 676 (D. Del. 2016). Because it does not carry that burden, I also grant the motion to exclude this portion of Cox’s opinion. * * * * * Expert testimony cannot be admitted if it is unreliable. I deny the parts of Ross’s motions to exclude opinions from Jonathan Krein and James Malackowski that were
as yet unresolved, as well as Plaintiffs’ motion to exclude Frederiksen-Cross’s opin- ions. But because Cox’s expert opinions on disgorgement and lost profits are not reli- able, I grant the portion of Plaintiffs’ motion excluding those opinions.