Pltfs Brief in Opp To Defs MSJ REDACTED

Case 1:20-cv-04160-JGK-OTW Document 169 Filed 09/02/22 Page 1 of 41
UNITED STATES DISTRICT COURT

SOUTHERN DISTRICT OF NEW YORK
X
HACHETTE BOOK GROUP, INC., :
HARPERCOLLINS PUBLISHERS LLC, JOHN : Case No. 1:20-cv-04160-JGK
WILEY & SONS, INC., and PENGUIN :
RANDOM HOUSE LLC, :
:
Plaintiffs, :
:
- against - :
:
INTERNET ARCHIVE and DOES 1 through 5, :
inclusive, :
:
Defendants. :
X
PLAINTIFFS’ MEMORANDUM OF LAW IN OPPOSITION TO INTERNET

ARCHIVE’S MOTION FOR SUMMARY JUDGMENT
DAVIS WRIGHT TREMAINE LLP

Elizabeth A. McNamara
Linda J. Steinman
John M. Browning
Jesse Feitel
Carl Mazurek
1251 Avenue of the Americas, 21st Floor
New York, NY 10020
Phone: (212) 489-8230
Email: lizmcnamara@dwt.com
lindasteinman@dwt.com
jackbrowning@dwt.com
jessefeitel@dwt.com
carlmazurek@dwt.com
OPPENHEIM + ZEBRAK, LLP

Scott A. Zebrak (pro hac vice)
Matthew J. Oppenheim
Danae Tinelli (pro hac vice)
4530 Wisconsin Avenue, NW, 5th Floor
Washington, DC 20016
Phone: (202) 480-2999
Email: matt@oandzlaw.com
scott@oandzlaw.com
danae@oandzlaw.com
Attorneys for Plaintiffs Hachette Book Group, Inc., HarperCollins Publishers LLC,
John Wiley & Sons, Inc. and Penguin Random House LLC
TABLE OF CONTENTS
Page
PRELIMINARY STATEMENT .................................................................................................... 1
ARGUMENT .................................................................................................................................. 2
I. CONTROLLED DIGITAL LENDING IS NOT A DEFENSE TO

COPYRIGHT INFRINGEMENT........................................................................... 2
A. Congress and the Copyright Office Have Repeatedly Considered

and Rejected a Digital First Sale Right ....................................................... 3
B. The Second Circuit and Other Courts Rejected the Premises of

Internet Archive’s Expansive CDL Theory ................................................ 4
1. The Owned-To-Loaned Ratio Does Not Save IA ........................... 4
2. ReDigi Rejected Internet Archive’s Argument that Fair Use

Can Expand the Statutory First Sale Doctrine ................................ 6
C. The Facts Disprove Internet Archive’s Presentation of CDL as a

Widespread and Lawful Practice ................................................................ 9
1. Internet Archive Grossly Inflates the Prevalence of CDL .............. 9
2. Internet Archive Does Not Obey CDL’s Core Principles ............. 11
II. FACTOR ONE WEIGHS IN FAVOR OF PLAINTIFFS .................................... 13
A. IA’s Infringement Is Not Transformative ................................................. 13
B. IA Ignores the Commercial Aspects of Its Website and Operations ........ 15
III. FACTOR TWO WEIGHS IN FAVOR OF PLAINTIFFS ................................... 16
IV. FACTOR THREE WEIGHS IN FAVOR OF PLAINTIFFS ............................... 17
V. FACTOR FOUR WEIGHS IN FAVOR OF PLAINTIFFS ................................. 18
A. Internet Archive Fails to Pay Fees that Libraries Customarily Pay

to Lend Plaintiffs’ Ebooks ........................................................................ 18
B. Undisputed Evidence Establishes that IA Offers A Competing

Substitute Resulting in Potential Lost Ebook Sales .................................. 20
1. IA’s Experts Do Not Rebut the Evidence of Market Harm .......... 25
i
2. Jorgensen’s Analysis is Unreliable ............................................... 26
3. Reimers Provides Limited and Unreliable Testimony .................. 28
VI. INTERNET ARCHIVE IS NOT ENTITLED TO REMITTANCE OF

STATUTORY DAMAGES UNDER SECTION 504 .......................................... 30
CONCLUSION ............................................................................................................................. 33
CERTIFICATION OF COMPLIANCE ....................................................................................... 35
ii
TABLE OF AUTHORITIES
Page(s)
Cases
A&M Records v. Napster,

239 F.3d 1004 (9th Cir. 2001) .................................................................................................14
Arista Records v. Lime Grp.,

2011 WL 1642434 (S.D.N.Y. Apr. 20, 2011)..........................................................................32
Association of Am. Publishers, Inc. v. Frosh,

2022 WL 484926 (D. Md. Feb. 16, 2022) .................................................................................3
Authors Guild, Inc. v. HathiTrust,

755 F.3d 87 (2d Cir. 2014).........................................................................................7, 8, 13, 17
Authors Guild v. Google,

804 F.3d 202 (2d Cir. 2015)...........................................................................................1, 15, 17
Blanch v. Koons,
467 F.3d 244 (2d Cir. 2006).....................................................................................................15
Bureau of Nat’l Literature v. Sells,

211 F. 379 (W.D. Wash. 1914) ..................................................................................................7
Campbell v. Acuff-Rose Music,

510 U.S. 569 (1994) .................................................................................................................13
Capitol Records, LLC v. ReDigi Inc.,

910 F.3d 649 (2d Cir. 2018)............................................................................................. passim
City of Providence v. Bats Global Markets,

2022 WL 902402 (S.D.N.Y. 2022) ..........................................................................................18
De Fontbrune v. Wofsy,
39 F.4th 1214 (9th Cir. 2022) ..................................................................................................16
Disney Enterprises v. VidAngel,

869 F.3d 848 (9th Cir. 2017) .................................................................................................5, 6
Doan v. American Book Co.,

105 F. 772 (7th Cir. 1901) .........................................................................................................7
Fox News Network v. TVEyes, Inc.,

883 F.3d 169 (2d Cir. 2018)...............................................................................................13, 17
iii
Harper & Row Publishers v. Nation Enters.,

471 U.S. 539 (1985) .................................................................................................................17
In re Electronic Books Antitrust Litigation,

2014 WL 1282298 (S.D.N.Y. 2014) ........................................................................................18
Infinity Broad. Corp. v. Kirkwood,

150 F.3d 104 (2d Cir. 1998).....................................................................................6, 13, 16, 23
Lamarr-Arruz v. CVS Pharmacy,

2017 WL 4277188 (S.D.N.Y. 2017) ........................................................................................18
On Davis v. The Gap,

246 F.3d 152 (2d Cir. 2001).....................................................................................................20
Oracle Corp. v. SAP AG,

765 F.3d 1081 (9th Cir. 2014) .................................................................................................20
Penguin Grp. (USA) v. Am. Buddha,

2015 WL 11170727 (D. Ariz. May 11, 2015) .........................................................................16
Ringgold v. Black Entm’t Television,

126 F.3d 70 (2d Cir. 1997).................................................................................................20, 21
Soc’y of Holy Transfiguration Monastery v. Gregory,

689 F.3d 29 (1st Cir. 2012) ................................................................................................16, 24
Sony Corp. of Am. v. Universal City Studios,

464 U.S. 417 (1984) .................................................................................................9, 13, 14, 17
Twin Peaks Prods., Inc. v. Publ’ns Int’l,

996 F.2d 1366 (2d Cir. 1993)...................................................................................................23
UMG Recordings v. MP3.com, Inc.,

92 F. Supp. 2d 349 (S.D.N.Y. 2000)..........................................................................................6
Worldwide Church of God v. Phila. Church of God,

227 F.3d 1110 (9th Cir. 2000) .................................................................................................17
Federal Statutes
17 U.S.C.
§ 106...........................................................................................................................................2
§ 107.....................................................................................................................................5, 22
§ 108.....................................................................................................................................8, 31
§ 109................................................................................................................................. passim
§ 504.......................................................................................................................30, 31, 32, 33
iv
Rules
F.R.E. 702 ......................................................................................................................................18
Other Authorities
144 Cong. Rec. H.7074 (1998) ........................................................................................................3
Dep’t of Commerce Internet Policy Task Force, White Paper on Remixes, First
Sale, and Statutory Damages 58 (2016) ....................................................................................4
S. Rep. 105-190 (1998) ..................................................................................................................31
U.S. Copyright Office, Library of Cong., Digital Millennium Copyright Act §104
Report vi (2001) .........................................................................................................................4
v
PRELIMINARY STATEMENT
Internet Archive concedes that it directly infringed the Works in Suit, which are but a
small sample representing its unauthorized scanning and distributing of millions of ebooks. IA’s
only defense rests on a manufactured doctrine, “controlled digital lending” (“CDL”). IA argues
CDL “is fundamentally the same as traditional library lending” and should be treated as a fair use
since it “furthers the ends of copyright’s exhaustion doctrine” (also known as the first sale
doctrine, codified at 17 U.S.C. §109). ECF No. 106 (“IA Br.”) at 1, 15. This position is a study
in blind denial that ignores established law and undisputed facts. Perhaps most astonishing, IA
essentially disregards ReDigi, in which the Second Circuit held that (1) unauthorized
reproduction is “not protected by §109”; (2) the fair use doctrine cannot be used to expand the
statutory scope of §109; and (3) ReDigi’s actions were unlawful even though it used technology
to avoid increasing the number of music files in circulation, a practice akin to CDL’s central
principle. Capitol Records, LLC v. ReDigi Inc., 910 F.3d 649, 659 (2d Cir. 2018). Nor does
Internet Archive even acknowledge the cases that prohibit format shifting in similar settings or
the Second Circuit’s warning that demonstrates that the Plaintiffs have a “strong” case against IA
for republishing entire unauthorized ebooks on a vast scale. Authors Guild v. Google, 804 F.3d
202, 226 (2d Cir. 2015); ECF No. 99 (“Pubs. Br.”) at 2-3.
Internet Archive’s motion is similarly disingenuous with the facts. It blithely ignores, for
instance, the commercial aspects of IA’s business (including its symbiotic relationship with
Better World Books (“BWB”)) and clear market harms felt by the Publishers and their authors,
including IA’s failure to pay millions in lost fees. In the end, IA displays contempt for authors,
like Sandra Cisneros, who rely on copyright to secure their livelihoods.
Ultimately, IA’s motion (buttressed by amicus briefs) amounts to a policy argument that
seeks to undercut the Copyright Act and Congressional policymakers, as well as reverse
1
controlling precedent. Such “a ruling … would exceed the proper exercise of the court’s
authority.” ReDigi, 910 F.3d at 664. IA cannot carry its burden to establish fair use, and its
motion for summary judgment should be denied, as should its motion for remittance of statutory
damages.
ARGUMENT
I. CONTROLLED DIGITAL LENDING IS NOT A DEFENSE TO COPYRIGHT

INFRINGEMENT
Internet Archive nominally moves for summary judgment on fair use grounds, but its
entire defense for “loaning … books over the Internet” is rooted in an expansive conception of
the first sale doctrine that the Second Circuit and Congress have explicitly rejected. As IA
tacitly admits, the Copyright Act’s first sale exception does not permit the creation, via
reproduction, of millions of unauthorized ebooks. To the contrary, §109(a) states that
“notwithstanding the provisions of section 106(3)” – the distribution right – “the owner of a
particular copy … is entitled, without the authority of the copyright owner, to sell or otherwise
dispose of the possession of that copy.” 17 U.S.C. §109 (emphasis added). IA likewise
acknowledges the Second Circuit’s holding in ReDigi “that because the statute addresses
distribution but not reproduction, Section 109(a) does not provide an independent defense to a
reproduction claim.” IA Br. at 20 n.3 (citing ReDigi, 910 F.3d at 664). In other words, IA
recognizes that the text of the Copyright Act does not contain a broad “exhaustion rule” that
permits its practice of CDL.
In an awkward attempt to sidestep these insurmountable hurdles, Internet Archive argues
that “CDL helps fulfill the goals of copyright’s exhaustion doctrine” and thus “the fair use
doctrine steps in to ensure that the substance of the common-law exhaustion rule survives
technological changes in the way copyright works are loaned and read.” IA Br. at 19-20
2
(emphasis added). This audacious argument fails for many reasons – not least because the
Second Circuit expressly rejected the idea that there is a hidden exhaustion doctrine that extends
the boundaries of §109.
A. Congress and the Copyright Office Have Repeatedly Considered and

Rejected a Digital First Sale Right
Having conceded that §109 does not protect CDL, Internet Archive asks this Court to
bypass the statutory text and interpose a digital exhaustion doctrine that, according to IA, was
accidentally omitted from the Copyright Act. The crux of IA’s argument is that codification of
§109 “predated the spread of copyrighted works” and thus inadvertently failed to account for
“the reproduction right that is implicated by digital lending.” IA Br. at 19. But IA and its amici
brazenly ignore that Congress repeatedly has considered and declined to amend §109 to
encompass digital works. ReDigi, 910 F.3d at 659; Association of Am. Publishers, Inc. v. Frosh,
2022 WL 484926, at *1 n.1 (D. Md. Feb. 16, 2022).
In 1998, the U.S. implemented the WIPO Internet Treaties through the Digital
Millennium Copyright Act. During DMCA proceedings, the concept of a “digital first sale”
doctrine was raised and rejected, including a provision that would have made the further
dissemination of digital files contingent on the deletion of the original copy after transmission.
The presiding Chairman explained that this would remove contractual controls placed by the
copyright holder and would “pos[e] so great a burden on a copyright owner so as to undermine
the incentive to create works which is the driving force behind the Copyright Act.”1
1
144 Cong. Rec. H.7074 (1998), https://www.congress.gov/105/crec/1998/08/04/CREC-1998-08-04-pt1-
PgH7074-3.pdf.
3
In 2001, at the request of Congress, the Register of Copyrights delivered a “multi-year
evaluation” on whether §109 should be updated due to “existing and emergent technology,”
which included a review of comments from the public, academia, libraries, copyright
organizations and copyright owners. ReDigi, 910 F.3d at 659 (citing U.S. Copyright Office,
Library of Cong., Digital Millennium Copyright Act §104 Report vi, 79-80 (2001)). The
Register concluded that §109 should not be amended and Congress left §109 unchanged. Id.
More recently, the House Judiciary Committee conducted a “comprehensive [ ] review”
of the Copyright Act, including a June 2014 hearing on digital first sale, and §109 again
remained unchanged. SUMF¶¶403-05. Thereafter, a 2016 White Paper by the Department of
Commerce’s Internet Policy Task Force recommended against expanding §109, finding that “the
risks to copyright owners’ primary markets as described by the Copyright Office in its 2001
Report do not appear to have diminished....” Dep’t of Commerce Internet Policy Task Force,
White Paper on Remixes, First Sale, and Statutory Damages 58 (2016),
https://www.uspto.gov/sites/default/files/documents/copyrightwhitepaper.pdf. The lack of a
digital first sale doctrine is thus no accident.
B. The Second Circuit and Other Courts Rejected the Premises of Internet
Archive’s Expansive CDL Theory
1. The Owned-To-Loaned Ratio Does Not Save IA
Internet Archive effectively asks this Court to embrace CDL by recognizing a hidden
“exhaustion” right beyond the text of the Copyright Act, but the courts have rejected analogous
efforts to bend copyright law to suit technology companies pushing theories similar to IA’s
“owned-to-loaned” ratio, the core of its CDL theory. The defendant in ReDigi, for instance,
developed a program that allowed users to “resell” legally-acquired digital music files through a
process intended to mimic the resale of physical phonorecords. The system deleted the original
4
file from the seller’s computer at the same time as a copy was made on ReDigi’s servers. 910
F.3d at 652-54. Echoing the core principle of CDL – that the physical book is never in use at the
same time as its digital copy – ReDigi argued its program ensured that each digital file never
existed in more than one place simultaneously. Id.
The Second Circuit rejected ReDigi’s first sale and fair use arguments, holding that the
measures ReDigi took to avoid increasing the total number of copies in existence did “not rebut
or nullify the fact that” ReDigi’s program unquestionably created new copies of each work,
violating the plaintiffs’ exclusive right of reproduction. Id. at 657-58. The Second Circuit’s
proscriptions apply with even greater force to CDL. While ReDigi involved digital-to-digital
duplication, CDL transgresses a step further by converting physical books into digital ebooks.
This is classic format shifting (sometimes called space-shifting), which is repeatedly rejected by
courts.
In Disney Enterprises v. VidAngel, 869 F.3d 848 (9th Cir. 2017), for example, the Ninth
Circuit forcefully rejected an analogous claim by an online streaming service that made movies
and television shows more “family friendly” by removing “objectional” content. The company
purchased “multiple authorized DVDs or Blu-ray discs” for each movie in its service, which it
then converted to a live streaming format that filtered out objectionable material. 869 F.3d at
853-54. In order to view a movie, a customer was required to temporarily “purchase” a disc,
which could be returned after viewing. Id. In other words, every movie streamed to a customer
was streamed against a lawfully-acquired copy on a disc. Nonetheless, the Ninth Circuit easily
rejected VidAngel’s first sale argument and its argument that such format-shifting constituted
fair use. 869 F.3d at 856-57, 860-62. As the Ninth Circuit declared: “The reported decisions
unanimously reject the view that space-shifting is fair use under §107.” Id. at 862. It
5
underscored the Register of Copyright’s conclusion that “[t]he law of fair use ... does not
sanction broad-based space-shifting or format-shifting.” Id.
UMG Recordings v. MP3.com, Inc., 92 F. Supp. 2d 349 (S.D.N.Y. 2000), similarly
involved an online music service that permitted subscribers to store and listen to music on
lawfully-acquired CDs via the Internet. In order to operate the service, the defendant purchased
tens of thousands of CDs and converted them to MP3 files. Subscribers were required to prove
that they owned the CD. Despite the veneer of a “lawfully acquired” foundation – similar to
CDL – Judge Rakoff found defendant’s fair use argument “indefensible.” He held that
“unauthorized copies ... being retransmitted in another medium [is] an insufficient basis for any
legitimate claim of transformation.” 92 F. Supp. 2d at 351. While MP3.com, like IA, tried to
argue that its service benefited the public, Judge Rakoff proclaimed that, “defendant’s ‘consumer
protection’ argument amounts to nothing more than a bald claim that defendant should be able to
misappropriate plaintiffs’ property simply because there is a consumer demand for it. This
hardly appeals to the conscience of equity.” Id. at 353; see also Infinity Broad. Corp. v.
Kirkwood, 150 F.3d 104, 108 & n.2 (2d Cir. 1998).
In sum, the case law ignored by IA teaches that the mere possession of a corresponding
physical copy does not magically make it lawful to engage in unauthorized reproduction and
distribution of the underlying work. IA does not have the unilateral right to “misappropriate
plaintiffs’ property” (id.) and reap the benefits of digital media with total disregard for the
Publishers’ carefully balanced licensing terms governing ebooks.
2. ReDigi Rejected Internet Archive’s Argument that Fair Use Can

Expand the Statutory First Sale Doctrine
Having acknowledged that §109 does not protect CDL, Internet Archive argues that the
fair use doctrine “steps in” to allow the “substance of the common-law exhaustion rule” to
6
inform the fair use analysis. IA Br. at 19-20. To further this argument, IA falsely asserts that the
“ReDigi court did not address the impact of the exhaustion doctrine on the fair use inquiry.” Id.
n.3. In reality, the Second Circuit not only rejected each argument IA puts forth, it also
specifically considered and rejected amicus briefs that IA and its amici in this action submitted in
ReDigi to make precisely the same exhaustion arguments they rehash here.
First, Internet Archive asserts that §109 does not “abrogate the common-law principle
favoring the ability of the owner of a copy to freely give, sell, or lend it.” Id. at 19-20. Then,
citing to a 1901 case, Doan v. American Book Co., 105 F. 772 (7th Cir. 1901), IA argues that the
“common-law doctrine of exhaustion can encompass reproduction of copyrighted material in
some circumstances.” Id. at 21. But the 120-year old Doan decision merely held that the first
sale doctrine allowed owners of a book to copy and replace its cover in order to “maintain the
book as nearly as possible in its original condition” – which is not remotely comparable to IA’s
mass digitization scheme. More fundamentally, ReDigi explicitly rejected the notion that a
common-law first sale doctrine expands the narrow limits of §109. To make this point, Judge
Leval cited a case that is materially indistinguishable from Doan – Bureau of Nat’l Literature v.
Sells, 211 F. 379 (W.D. Wash. 1914) – as an example of a common law decision that did not
survive §109. ReDigi, 910 F.3d at 664 (“Congress, in promulgating §109(a), adopted a narrower
conception [of first sale].”).
Second, Internet Archive argues that CDL is a favored use under the fair use inquiry
because it “advances the purposes of narrow statutory exceptions,” namely the concededly
inapplicable first sale doctrine. IA Br. at 20. In support of this point, IA cites to Authors Guild,
Inc. v. HathiTrust, where the Second Circuit held that a program making books available to the
blind was a favored fair use, even if it did not satisfy the strict language of the Chafee
7
Amendment permitting certain copying for the blind. 755 F.3d 87, 102 (2d Cir. 2014). Yet, IA
made the exact same argument to the Second Circuit in the amicus brief it (and other amici)
submitted in ReDigi, which contended that ReDigi’s service furthered the general purpose of
§109 and was thus a favored use under HathiTrust. See ECF No. 96 (“McN Decl.”) Ex. 137
(“IA ReDigi Br.”) at 5-16. Judge Leval firmly rejected this argument, and declined to adopt IA’s
conception of the animating spirit of first sale in order to create a “fundamental entitlement,
without regard to the terms of §109(a).” 910 F.3d at 664. As he explained, “such a ruling would
exceed the proper exercise of the court’s authority” since §109 is a statutory provision in which
Congress “has taken control, dictating both policy and the details of its execution.” 910 F.3d at
664.2 In short, fair use is not a vehicle to expand §109 contrary to the will of Congress.
A decision recognizing Internet Archive’s practice of CDL as fair use also would
undermine the Congressional intent behind the limited exceptions that permit libraries to copy
lost or damaged books under §108. Despite claiming to be a library, IA notably includes no
mention of §108 – understandably because the systematic creation and distribution of digital
copies of millions of books readily available for authorized acquisition is far removed from the
provision’s careful boundaries. Yet, together with the first sale doctrine, these are the very
sections of the Copyright Act that primarily effectuate Congress’ “intent to secure libraries’ right
to lend books they own.” IA Br. at 20.
Fundamentally, what IA seeks is not application of the law as it stands, but rather for the
Court to fashion a defense to infringement that does not currently exist. To this end, IA and its
2
IA’s own outside counsel wrote that ReDigi “raises questions concerning the viability of Controlled
Digital Lending ... by libraries.” SUMF¶¶456-60.
8
amici trumpet the supposed “magnitude of the public benefit that flows from CDL.”3 IA Br. at 7,
24. Again, ReDigi foreclosed this policy-first approach. As Judge Leval observed, courts are
“poorly equipped to assess” the broad implications of copyright policy or weigh its alleged
benefits against drawbacks. ReDigi, 910 F.3d at 664; see also Sony Corp. of Am. v. Universal
City Studios, 464 U.S. 417, 431 (1984) (acknowledging the “recurring theme” of the “judiciary’s
reluctance to expand the protections afforded by . . . copyright without explicit legislative
guidance”). Respectfully, we urge this Court to “reject the invitation to substitute [its] judgment
for that of Congress.” ReDigi, 910 F.3d at 664.
C. The Facts Disprove Internet Archive’s Presentation of CDL as a Widespread

and Lawful Practice
1. Internet Archive Grossly Inflates the Prevalence of CDL
Having failed to establish that CDL is permissible under copyright law, Internet Archive
and its amici present CDL as a fait accompli, contending that “[l]ibraries have been practicing
CDL in one form or another for more than a decade, and hundreds of libraries use it to lend
books digitally today.” IA Br. at 1; see also ECF No. 146 (“Wu Br.”) at 11-14; ECF No. 142
(“Library Futures Br.”) at 3, 7-9.
Yet the evidence proves that the current version of CDL is a recent phenomenon. While
IA began “lending” public domain or out-of-print books on a small scale from about 2011, it
only distributed them on library networks or within limited geographical areas. SUMF¶¶224,
3
The amicus briefs eschew legal argument and are filled instead with thinly supported prescriptions for
public policy that courts are ill-equipped to assess, with only three of the six briefs even bothering to cite
ReDigi or argue CDL is permitted under copyright law. Further, IA amici are largely barely-disguised
surrogates of IA with whom it collaborated extensively on manufacturing the legal theories at issue in this
litigation. See Plaintiffs’ Opposition to Requests for Leave to Appear as Amici Curiae (ECF No. 138).
9
227, 352. Over the years, many libraries were reluctant to allow IA to scan in-copyright books
because of “copyright concerns.” Id. ¶¶297-299. IA’s digitization and lending capacity only
began to grow exponentially in 2018-19 with the purchase of BWB and the launch of IA’s Open
Libraries project. Id. ¶¶232-233, 333, 355, 338. Even so, only 62 Partner Libraries contributed
books to the Open Libraries project out of the 9,000 library systems in the United States,
encompassing some 15,000 branches, with a mere 13 of those being public libraries. Simply put,
there is no evidence that CDL is a longstanding and widespread practice – particularly with
respect to public libraries, only a handful of which endorsed IA’s statement on CDL. Id. ¶¶392,
462.
The only evidence IA and its amici identify to support their grandiose contention of
widespread CDL adoption is a few academic programs. See IA Br. at 31-32; Library Futures Br.
at 7-8.4 But closer inspection reveals that these programs are all highly restricted. Each of the
CDL programs is limited to lending from collections of university libraries and providing access
only to members of those universities; none involve a library making its collection freely
available to the world. Further, the programs were implemented not in the regular course of
4
Contrary to Library Futures’ brief, the Boston Library Consortium of academic libraries appears to be
only studying CDL at this stage. See Boston Library Consortium to implement controlled digital lending
for interlibrary loan, BOSTON LIBRARY CONSORTIUM, https://blc.org/news/120355.
10
business, but in response to access issues precipitated by COVID.5 Even then, a number of these
programs provided access only to very limited works within the libraries’ collections.6
2. Internet Archive Does Not Obey CDL’s Core Principles
Even if CDL somehow were fair use – and it is not – Internet Archive’s actions are still
infringement because it defies the most basic principle of its own theory – i.e., the “owned-to-
loaned” ratio. IA argues, at bottom, that its use of technical controls to “enforce a one-to-one
owned-to-loaned ratio, is fair use” because this ensures that the user who reads an ebook on the
Website “is the one person in the world who is then borrowing that particular, bought-and-paid-
for library book.” IA. Br. at 1, 7, 14, 18.
The record belies these claims, beginning with the assertion that IA merely “lend[s] its
books to its patrons.” Id. at 1, 4 (emphasis added). Starting in 2018, and increasingly since then,
IA has assumed the mantle of a national library that lends books owned by other libraries.
Specifically, IA “partners” with libraries nationwide under the Open Libraries project, runs
“overlap analysis” to identify books that IA scanned already, and counts those books from
partner collections towards the number of ebooks available for concurrent loans on the Website.
Pubs. Br. at 15-16; IA Br. at 7. In effect, IA adopts the “royal we” by counting Partner Libraries’
books towards its own “owned-to-loaned” ratio.
5
HathiTrust and a handful of university libraries, for instance, first enacted CDL programs specifically in
response to the pandemic and tightly conditioned these programs on library closures. See Supplemental
Declaration of Elizabeth A. McNamara (“Suppl. McN Decl.”), ¶12.

6
Carnegie Mellon University and Michigan State University offered 60 titles and 206 course-related
titles, respectively, for CDL. See id.
11
Further, IA cannot claim, as it does, that lending limits are “backstopped by strong
technical protections” after admitting it has “no way of knowing ... whether the physical and
digital copies of a book [are] in circulation simultaneously.” IA Br. at 1; SUMF¶495. Far from
verifying whether Partner Libraries have taken physical copies out of circulation, IA even
concedes its awareness that certain Partner Libraries do not remove the physical books from their
shelves. SUMF¶494; ¶¶504-506. The upshot is that IA cannot ensure it abides by its own
generous conception of the owned-to-loaned principle. SUMF¶479.
Finally, IA abandoned any façade of fair use when it removed the pretense of the owned-
to-loaned ratio and set aside all lending limits to implement the grandly-titled National
Emergency Library. SUMF¶¶542-43; see also ECF No. 90-9 (Foster Decl. Ex. 6) at 4. IA
makes the remarkable argument that even this use complied with the spirit of the owned-to-
loaned principle because nationwide library closures “took 650 million books out of circulation”
and “there were surely more non-circulating copies locked up in shuttered libraries than would
be borrowed via the Internet Archive.” IA Br. at 9. In other words, IA unilaterally conferred on
itself the right to lend against every physical library book in the United States. IA never explains
who granted it such a lofty status. Surely, neither Congress nor the Copyright Office, which
frowned on its practices, designated IA the “national library” of the United States. SUMF¶¶551-
554.
This presumptuous argument represents the logical end point of the Open Libraries
project: to enable Internet Archive to appoint itself the de facto national (or global) digital
library with the power to distribute as many ebooks simultaneously as there are physical books
on the shelves of the 9,000 library systems throughout the country and those beyond its shores
(or at least as many libraries that become Partner Libraries if the Court greenlights CDL). This
12
makes a mockery of the pretense that CDL is a “controlled” method for distributing ebooks. The
Court should declare it to be the blatant infringement that it is.
II. FACTOR ONE WEIGHS IN FAVOR OF PLAINTIFFS
A. IA’s Infringement Is Not Transformative
No other rationale offered by Internet Archive supports fair use. The central problem
with IA’s fair use defense is that there is absolutely nothing transformative about simply
“repackag[ing] [and] republish[ing]” the full text of the Publishers’ books in unauthorized digital
formats and disseminating them worldwide. HathiTrust, 755 F.3d at 96; see also, e.g., Fox News
Network v. TVEyes, Inc., 883 F.3d 169, 177 (2d Cir. 2018); Infinity, 150 F.3d at 108. IA fails to
cite the Supreme Court’s definition of transformativeness, presumably because it destroys its
argument. The core focus is “whether the new work supersede[s] the objects of the original
creation … or instead adds something new, with a further purpose or different character, altering
the first with new expression, meaning or message.” Campbell v. Acuff-Rose Music, 510 U.S.
569, 579 (1994); see also ReDigi, 910 F.3d at 660-61. Here, IA’s ebooks add nothing new and
are distributed to be read, which serves the same purpose as the original.
Internet Archive relies on Sony, and the discussion of Sony in TVEyes and ReDigi, to
argue that it has expanded the utility of the Works, but it cannot shoehorn CDL into that very
different setting. Sony involved “the private, home use of VCRs for recording programs
broadcast on the public airwaves without charge”; “the primary use of the [Betamax] machine
for most owners was ‘time-shifting’—the practice of recording a program to view it once at a
later time, and thereafter erasing it.” Sony, 464 U.S. at 423. As the Supreme Court noted, the
case posed “no issue concerning the transfer of tapes to other persons, the use of home-recorded
tapes for public performances, or the copying of programs transmitted on pay or cable television
systems.” Id. at 425.
13
In stark contrast, IA is an organization systematically republishing millions of
unauthorized, in-copyright ebooks and loaning them out 25 million times a year to the general
public. Thus, while Sony posed “no issues concerning the transfer of tapes to other persons,”
IA’s whole purpose is “the transfer of [ebooks] to other persons.” Id. Thus Sony is
distinguishable, as it was in Napster, because the time-shifting in Sony “exposed the material
only to the original user” and “did not also simultaneously involve distribution of the
copyrighted material to the general public.” A&M Records v. Napster, 239 F.3d 1004, 1019 (9th
Cir. 2001).
Further, unlike television channels broadcasting free content to anyone with a television
set, the Publishers require payment from purchasers of print books and authorized ebooks – and
thus this is not a case where Internet Archive is “entitled to receive the content” it
misappropriates. ReDigi, 910 F.3d at 661. Sony specifically noted that that case did not involve
“the copying of programs transmitted on pay or cable television systems.” 464 U.S. at 425. No
court has ever suggested that the limited right to record broadcast television for private
consumption can justify the systematic scanning and worldwide distribution of millions of
ebooks that are already available via library licensing programs, with fees and controls tailored
to that market. ECF No. 95 (“Weber Decl.”) ¶70; ReDigi, 910 F.3d at 661.
Internet Archive also tries to squeeze into Sony by asserting that “CDL makes the
delivery more efficient than physically handing or mailing that same print book” (IA Br. at 17),
but this misses the crucial point that IA’s ebooks are no more “efficient” than the millions of
authorized ebooks that libraries already license. This argument only underscores that ebooks are
distinct products with advantages that print books cannot match, which is precisely why the
Publishers invested tens of millions of dollars to develop digital publishing technologies,
14
infrastructure, licensing models and pricing structures for ebooks in retail and library markets.
Pubs. Br. at 6-9.
Finally, Internet Archive suggests that Wikipedia links to its ebooks render its entire CDL
program fair use. But this is not the primary function of IA’s Website and this peripheral use
does not justify the distribution of entire ebooks.7 Moreover, IA’s Wikipedia links are not like
the snippets in Google Books because anyone who follows a link to IA’s Website can read the
full book.
In sum, IA fails the “heart of the fair use inquiry” because CDL is not transformative.
Blanch v. Koons, 467 F.3d 244, 251 (2d Cir. 2006).
B. IA Ignores the Commercial Aspects of Its Website and Operations
Internet Archive also argues unconvincingly that its practice of “CDL is wholly
noncommercial” because IA is “a 501(c)(3) public charity lending books for personal and
noncommercial reading by individual members of the public.” IA Br. 16. This conclusion
stunningly ignores the deep commercial entanglements between IA and BWB, a for-profit
company effectively owned and operated by Brewster Kahle.
As a threshold matter, the Second Circuit has held that “there is … no reason to presume
categorically that a nonprofit educational purpose should qualify as a fair use.” Google Books,
804 F.3d at 219 n.20. That would give carte blanche to more than a million tax exempt entities
to infringe copyrighted material. More fundamentally, IA’s global distribution of millions of
general interest books, from romances to thrillers, for anyone to read for any reason does not
serve the type of “educational purpose” recognized under the Copyright Act, such as teaching.
7
This argument is so extraneous that IA did not even mention it in its interrogatory responses explaining
the basis for its fair use defense. Suppl. McN Decl. Ex 1 at 19-22.
15
Penguin Grp. (USA) v. Am. Buddha, 2015 WL 11170727, at *4-5 (D. Ariz. May 11, 2015). See
Pubs. Br. at 25-26.
Next, Internet Archive takes the indefensible position that its activities are “not
commercial in nature” because it “does not charge patrons any fees.” IA Br. 16. But the
commerciality analysis turns on whether the infringer “stands to gain recognition and financial
support through its posting of the works online.” Am. Buddha, 2015 WL 11170727, at *4; see
also Soc’y of Holy Transfiguration Monastery v. Gregory, 689 F.3d 29, 61 (1st Cir. 2012). Here,
it is undisputed that IA “monetized” every webpage on which the Works appear, including by
featuring buttons that invite users to “Purchase” copies from its affiliate BWB. Pubs. Br. 26-27.
Nor can IA whitewash the many other manifestly commercial elements of its enterprise –
including its multi-million dollar book scanning business, its effective purchase of BWB, the
world’s biggest for-profit used book store, to secure a “book pipeline” and the “synergies”
developed with BWB to channel “funding back to the Internet Archive to ensure that it can
continue to deliver free services to the world.” Id. at 27.
Last, Internet Archive asserts in conclusory fashion that its “[p]atrons’ borrowing and
private reading is … noncommercial in nature” (IA Br. at 16), but the law is clear that the “end-
user’s utilization of the product is largely irrelevant” to the analysis. De Fontbrune v. Wofsy, 39
F.4th 1214, 1224 (9th Cir. 2022); Infinity, 150 F.3d at 109 n.3.
III. FACTOR TWO WEIGHS IN FAVOR OF PLAINTIFFS
Internet Archive argues that the second factor is neutral because the 127 Works in Suit
“are of many different types,” but this grab-bag approach ignores that the Works include seminal
books of American literature, from Elie Wiesel to Malcolm Gladwell to Toni Morrison. IA Br.
at 22. All of the Works are fiction and non-fiction books containing “subjective descriptions and
portraits … whose power lies in the author’s individualized expression,” and should receive
16
maximum protection under the second factor. Harper & Row Publishers v. Nation Enters., 471
U.S. 539, 563-64 (1985); see also Worldwide Church of God v. Phila. Church of God, 227 F.3d
1110, 1118 (9th Cir. 2000). Moreover, contrary to IA’s contention, Google Books does not
support its position. There, the Second Circuit held the second factor to be neutral “not because
Plaintiffs’ works are factual, but because the secondary use transformatively provides valuable
information about the original.” 804 F.3d at 220. Here, IA’s use is not transformative.
Finally, Internet Archive’s claims to limit its Website to books published more than five
years ago do not aid its cause. IA Br. at 22-23. For one, this is an arbitrary, self-imposed rule
which IA intermittently violates and could change at any time. See McN Decl. ¶35. More
critically, IA’s justification totally ignores that the Publishers rely – and are entitled to rely – on
the critical revenue generated by backlist book sales over the full copyright term, which produce
vital royalties for authors and steady income that enables the Publishers to assume the risk of
publishing new works. SUMF¶¶32-37.
IV. FACTOR THREE WEIGHS IN FAVOR OF PLAINTIFFS
The third factor weighs heavily against fair use because IA does not rebut the heavy
presumption that wholesale copying clearly “tends to disfavor a finding of fair use.” ReDigi, 910
F.3d at 662; see also TVEyes, 883 F.3d at 179. IA claims that its copying aligns with Sony,
HathiTrust and Google Books, where the “copying of the entire copyrighted work was necessary
to effectuate the favored purpose and character of the defendant’s use.” IA Br. at 23. But IA’s
global ebook distribution scheme is far removed from the limited “favored purposes” found in
those actions – such as recording free television broadcasts at home, enabling accredited blind
users to access texts, or operating a search engine for books displaying short snippets. The third
factor cannot be neutral given IA’s avowed purpose, which is to allow anybody in the world with
an Internet collection to read every book ever published for free.
17
V. FACTOR FOUR WEIGHS IN FAVOR OF PLAINTIFFS
The fourth factor weighs heavily in favor of the Publishers. IA sidesteps the damning
record evidence against it and brazenly misrepresents that “the publishers have not offered any
evidence that Internet Archive’s digital lending, or anyone else’s, has cost them one penny in
revenue.” IA Br. at 2. And IA wrongly implies both that the burden of proof is on the
Publishers and that the relevant inquiry is to prove special damages, not market harm, while at
the same time ignoring that market harm need only be “potential” and can result from scaled-up,
widespread conduct. See Pubs. Br. at 20, 29-41. Finally, IA grossly overstates the limited and
unreliable conclusions of its experts, which should be disregarded by the Court. See Declaration
of Jeffrey T. Prince (“Prince Decl.”), passim.8
A. Internet Archive Fails to Pay Fees that Libraries Customarily Pay to Lend
Plaintiffs’ Ebooks
The Publishers have put forth two types of market harm. The first is that Internet
Archive, which holds itself out as a library, distributed ebooks on short term loans without
paying the Publishers the licensing fees that they receive from the existing market for library
ebooks. Pubs. Br. at 30-33; Suppl. McN Decl. Exs. 2 (“R-A Tr.”) at 141:3-141:11, 142:4-
8
In the event of a trial, the Publishers will move under F.R.E. 702 to exclude Rasmus Jorgensen’s and
Imke Reimers’ testimony, as well as select portions of Susan Hildreth’s opinions, since they are
unreliable and based on insufficient facts and impermissibly flawed principles and methods (which also
have not been reliably applied), including for all the reasons set forth herein and in the Prince Declaration,
including the failure to adequately control for other factors. City of Providence v. Bats Global Markets,
2022 WL 902402, at *12-13 (S.D.N.Y. 2022); Lamarr-Arruz v. CVS Pharmacy, 2017 WL 4277188, at
*10 (S.D.N.Y. 2017); In re Electronic Books Antitrust Litigation, 2014 WL 1282298 at *10-11 (S.D.N.Y.
2014).
18
142:10, 144:2-144:12; 3 (“Weber Tr.”) at 202:23-203:11, 205:22-206:9, 207, 214; 4 (“Pavese
Tr.”) at 51-54, 191-92; 6 (“Lazarus Tr.”) at 48-51; Prince Decl. Ex. 5 (“Hildreth Tr.”) at 47:11-
48:7. The Publishers have already lost out on millions of dollars in lost fees from IA.
SUMF¶586. If IA is allowed to scale up its operations, the Publishers would lose additional
significant revenues. SUMF¶¶575-88; Prince Decl. ¶¶5-8. This market harm is patently
obvious.
In its defense, Internet Archive suggests that somebody paid the Publishers for a print
edition of each Work and that is sufficient. IA Br. at 1, 30. But Publishers “do not price print
books with the expectation that they will serve as both print books and ebooks readily capable of
free worldwide distribution” (ECF No. 94 (“R-A Decl.”) ¶2), and it is elementary copyright law
that copyrights are divisible, and rightsholders are entitled to the proceeds of all formats, which
in turn incentivizes investment in new formats. See Section I(B)(1), supra.9
Next, IA argues that although it invites users to “Read Free Library Books Online,” the
Court should ignore the Publishers’ “traditional, reasonable, or likely to be developed” library
ebook market by which they receive millions in license fees. See Pubs. Br. at 29-39. Instead, IA
asserts the relevant market is CDL, and “[t]here is no traditional market for publishers to license
borrowing via CDL.” IA Br. at 32. This smoke and mirrors does not work. Of course there is
no “market for publishers to license borrowing via CDL” since the whole theory of CDL is
9
IA’s expert Reimers herself stated in a publicly posted draft article that, “[m]any publishers are opposed
to mass digitization projects .… This is a reasonable reaction, given that the mass digitization projects
typically digitize works without the publisher’s consent and without explicit licensing fees that would
generate direct revenue.” Suppl. McN Decl. Ex. 5 (emphasis added) at 17; see also Prince Decl. Ex. 4
(“Reimers Tr.”) at 102-105.
19
predicated on the assumption that no license is necessary if IA or its library partners own the
same book in a physical format. No defendant can define the relevant market as its own form of
infringement. If the Court accepted that untenable proposition, any infringer could tip the fourth
factor in their favor merely by claiming the market for infringing products is distinct from the
market for the legal product.
Finally, IA’s suggestion that there is no licensing market because IA refuses to license
ebooks conflicts with the well-accepted premise that a copyright owner can recover damages
whether or not the infringer would have agreed to the owner’s terms. Oracle Corp. v. SAP AG,
765 F.3d 1081, 1088 (9th Cir. 2014); On Davis v. The Gap, 246 F.3d 152, 165 (2d Cir. 2001). In
sum, this form of market harm alone suffices to tilt the fourth factor heavily in the Publishers’
favor. Ringgold v. Black Entm’t Television, 126 F.3d 70, 81 & n.16 (2d Cir. 1997).
B. Undisputed Evidence Establishes that IA Offers A Competing Substitute

Resulting in Potential Lost Ebook Sales
Internet Archive also harms the Publishers by offering a competing substitute for
authorized ebooks, leading to potential lost revenues from both public and academic libraries and
consumers.10 IA principally responds that, “[i]n a copyright lawsuit against a practice that has
continued for years, one would expect the copyright holder to be able to point to some metric
showing that the defendant’s conduct harmed them.” IA Br. at 25. Once again, IA ignores the
evidence showing the fallacy of this contention.
10
See SUMF¶¶255, 378-88, 617-20; R-A Tr. at 141-43; Weber Tr. at 202-07, 210-15; Pavese Tr. at 51-
54. IA’s selective quotation of out-of-context snippets from Publisher testimony distorts it. See, e.g.,
Weber Decl. Ex. 26 (“Dye Tr.”); Lazarus Tr. at 36-41.
20
First, there is indisputable evidence of market harm that is more powerful than the
admittedly “weak” and unreliable statistical evidence produced by IA’s experts. ECF No. 109-1
(“Reimers Report”) ¶¶38, 54. IA distributes the full ebook, often for a 14-day loan. The record
establishes that IA users frequently read these ebooks (SUMF ¶¶256-62; 236-69; 259-60) –
providing input like, “I was reading Daniel Silva’s Gabriel Allon [thriller] series on [IA].” McN
Decl. Ex. 14 and SUMF¶¶259-60. And IA identifies itself as competing for eyeballs with
Amazon and Overdrive. SUMF¶524.11 Further, as the Second Circuit has recognized, the “basic
market [for books] is the one-time reader.” Ringgold, 126 F.2d at 79. IA’s harm is compounded
by the fact that it has greatly increased the number of titles it offers from each of the Publishers
since it removed the Works in Suit in June 2020, now distributing a total of 33,000 works
(SUMF¶¶240, 522); for Penguin Random House, this is a stunning 48.4% of titles in its catalog
first published through 2015. SUMF¶583.f.
Moreover, the evidence is particularly strong with respect to library ebooks. IA has
launched campaigns urging public and academic libraries to use its Website instead of licensing
authorized ebooks. SUMF¶¶379-388. As IA’s Open Libraries agreement explicitly states, its
mission is to: “[M]ak[e] every library’s collection into a digital collection by allowing a library
to lend a digital version of physical volumes they own.” SUMF¶603. IA cannot plausibly argue
that its actions will have no effect on library ebook revenues when it engages in a systematic
effort to convince libraries to use the Open Libraries project as a substitute for paying for
authorized ebooks. In short, after launching its multi-year campaign, IA cannot say to this Court,
“No, really, we didn’t mean it.” Moreover, many libraries (both Partner Libraries and
11
For example, the record shows a school was negotiating a classroom license with PRH for two ebook
titles when the school informed OverDrive that it would instead use IA’s Website. SUMF¶¶563-64.
21
unaffiliated libraries) have begun incorporating links to the IA Website into their online library
catalogs or library websites.12
It is basic economic theory, and self-evident, that “when a free (zero-priced) competing
product enters the market, economists generally expect to observe a shift in consumer demand
toward the free good and thus a downward shift in demand for existing goods.” Prince Decl.
¶¶10-11. Both sides also agree that, as Reimers (IA’s expert) has written, “[r]ecent academic
studies show that illegal distribution displaces legal sales in media industries,” and that “most
recent work finds that legal sales are significantly displaced by pirated versions,” with one study
by Reimers showing an increase in sales by at least 14 percent on average following anti-piracy
measures.13 In that study, as in this case, the infringing copies were largely scans of print books,
not hacked ebook files, and featured backlist titles.14 While Reimers has tried to distance her
earlier findings from this matter, she relies on pure speculation to do so.15 Given the academic
research, it would require “a considerable amount of reliable evidence” to claim that IA’s
Website does not displace sales. Prince Decl. ¶14. As detailed below, IA and its experts provide
nothing of the kind. Id. at ¶¶14, 24-86.
Second, §107 looks not at actual past harm but at “the effect of the use upon the potential
market for or value of the copyrighted work.” As detailed in the Publishers’ moving brief (Pubs.
Br. at 33-36), IA distributes the Publishers “copyrighted material in a market that [the
12
See SUMF¶¶393-396; Prince Decl. ¶16.
13
Prince Decl. Ex. 6 (“Reimers 2016”), at 412-14, 421, 433; see also Reimers Tr. at 56:2-24, 58:11-21.
See also Prince Decl. Ex. 1 (“Prince Report”) ¶¶30-38.

14
Reimer 2016 at 415-16; Reimers Tr. at 45-46, 50.
15
See Reimers Tr. at 71:25 – 78:22.
22
Publishers], as the copyright owner, [are] exclusively entitled to exploit.… [Defendant] replaces
[Plaintiffs] as the supplier of those [ebooks] to meet the demand of [its] customers. This is
precisely the kind of harm the fourth factor aims to prevent.” Infinity, 150 F.3d at 111; Twin
Peaks Prods., Inc. v. Publ’ns Int’l, 996 F.2d 1366, 1377 (2d Cir. 1993) (fourth factor focuses on
whether the Defendant “competes in markets in which [the Publishers have] a legitimate
interest”).
Third, given the nature of the book market and the relative volume of IA loans, it would
be exceedingly difficult to create a reliable “metric” that isolates lost sales attributable to Internet
Archive and tunes out all the other factors impacting sales figures for an individual Work at any
given point in time. Prince Decl. ¶85. The difficulty of this Sisyphean task is underscored by
the striking failure of IA’s experts to adequately control for relevant factors. Id. ¶¶35-86. As the
Sevier Declaration explains, “[o]ur long term experience teaches us that each title is unique, and
there are myriad reasons why its sales or library checkouts in any particular format may go up or
down over two quarters in a particular year. In most circumstances this makes it difficult, if not
impossible, to isolate the impact of any one factor among the hundreds impacting a title’s sales
and checkout figures at any given point.” ECF No. 92 (“Sevier Decl.”) ¶80; see also Weber Decl.
¶88.
Moreover, as IA acknowledges, “[t]he lending numbers from the Internet Archive for the
Works in Suit” – which were all removed by mid-2020 – “are small relative to the size of the
library market for ebooks.” IA Br. at 25; see also ECF No. 109-2 (“Reimers Reply Report”) ¶19
(suggesting that before 2020, “the Internet Archive might have been too small to cause a sales
decrease”). While this does not mean there was no material harm from IA’s 57,000 unauthorized
loans, it does make it harder to create an accurate metric measuring that harm. Prince Decl.
23
¶¶83-85. However, in the two years since the Works were removed, IA has seen a meteoric rise
in members, monthly loans, and concurrent copies available for loan under the Open Libraries
project. SUMF¶¶247-50, 351-99, 544-45, 609; see also id. ¶546 (showing eightfold increase of
books borrowed over 28-day period between March 2020 and July 2022).
Given these escalating numbers, it is no surprise that Internet Archive fails to adequately
address the harm to the Publishers if its conduct or similar conduct were to become widespread.
A rightsholder is not required to “delay taking action to defend its copyright until actual (and
perhaps irreparable) harm had accrued.” Gregory, 689 F.3d at 64 n.23.
Internet Archive tries to downplay the future harm of ubiquitous CDL by arguing that if
the Publishers experienced no quantifiable reduction in library checkouts or consumer revenues
during the short NEL period, when IA distributed unlimited copies of the Works in Suit, then the
Publishers will suffer no harm if CDL “became widespread.” IA Br. at 13. Even accepting IA’s
premise for the sake of argument, this is an indefensible conclusion.
“Widespread conduct” includes the scenario in which Internet Archive succeeds in its
quest to become a central repository of ebooks – in effect, a national or global digital library –
and becomes much better known to the public. See SUMF¶¶232-35, 355-61. If the Court were
to condone its activities, IA could substantially expand the number of Partner Libraries in the
Open Libraries project from 65 currently to thousands. Not only could thousands of new Partner
Libraries contribute “concurrent copies” to increase the loan count, they and other libraries
would be able to link from their website card catalogs as a viable alternative to purchasing
authorized ebook licenses from the Publishers. Prince Report ¶¶ 66, 68-69 & nn.169, 170, 194;
SUMF¶¶378-96. As libraries around the country increasingly relied on IA to lend millions of
backlist titles in digital form, the loss of revenues from these public and academic libraries, as
24
well as from consumers, would create substantial market harm to publishers and authors. Prince
Decl. ¶¶76-80; SUMF¶608-612, 620. Likewise, other organizations could opportunistically
brand themselves a library and harm Plaintiffs by performing similar functions. This scenario is
far removed from the brief NEL period in Spring 2020. “At the extreme end of the spectrum,
there would be a scenario in which IA were known to all libraries and all consumers and had free
unauthorized digital books of all published backlist titles, with a sufficient number of copies to
meet worldwide demand without wait lists.” Prince Decl. ¶78. This would “destabilize [the
book] industry” – and other media industries vulnerable to the same scheme. Weber Decl. ¶¶92-
93.
1. IA’s Experts Do Not Rebut the Evidence of Market Harm
Nothing that IA’s experts state overcomes this evidence of market harm. First, none of
IA’s three experts perform an analysis of whether the Publishers lost revenues from public and
academic libraries because IA supplanted their authorized library ebook market. Reimers
focuses solely on the consumer print book market. Hildreth states that she does not have an
opinion on whether IA caused Plaintiffs economic harm. (ECF No. 110-2 (“Hildreth Reply
Report”) ¶7). Meanwhile, she concluded that “libraries ‘would reallocate their spending away
from ebook licensing for those titles being accessed through CDL and would spend more on
ebook licensing for newer titles or on print books.’” IA Br. at 33. This is a tacit admission that
publishers and authors affected by this “reallocation” necessarily lose revenue. Indeed, even
Hildreth conceded that libraries participating in CDL could reallocate money spent on library
ebooks to movies, CDs, and video games. Hildreth Tr. at 195:25-203:25; see also id. at 216:19-
218:18; Prince Decl. ¶¶26-28. Similarly, IA’s theory that CDL does not harm incentives to
create because it merely shifts library ebook expenditures from backlist books to new books
completely misunderstands that backlist revenue serves a critical purpose for publishers distinct
25
from frontlist revenue, and that many authors rely on a steady income over time. SUMF¶¶38-62;
ECF No. 89 (“Cisneros Decl.”) ¶9.
As for Jorgensen, he merely reviewed Overdrive check-outs of the Works during Q2 and
Q3 of 2020. Library check-outs do not equate with the impact of IA’s conduct on library ebook
licensing revenue because of the complex relationship between checkouts and revenues from
library ebook licenses – which Jorgensen admits. ECF No. 108-2 (“Jorgensen Reply Report”)
¶¶26-27; Prince Decl. ¶¶31-33. Most ebook licenses are in effect for two years or 26 circulations
and licenses are based on annual library budgetary cycles, so check-outs for two quarters would
say little about license revenues. Prince Decl. ¶31; Prince Decl. Ex. 3 (“Jorgensen Tr.”) at 101.
Notably, although Overdrive revenue data was produced in this matter,16 Dr. Jorgensen chose not
to analyze it.
2. Jorgensen’s Analysis is Unreliable
While not on point, Jorgensen’s analysis of Overdrive check-outs should be disregarded
by the Court as unreliable, as should his analysis of consumer ebook and paperback sales for the
fewer than 30 Hachette titles that are Works in Suit. Jorgensen starts from the premise that the
Works were removed from IA’s Website in June 2020, which coincided with the end of the NEL,
and then makes a grand leap, concluding that because on average the Overdrive check-out data
and Hachette sales figures for the Works happened to go down (although, as he admits, with
“considerable variation”17) between Q2 and Q3 of 2020 – and declined more than industry-wide
figures for the entire market – the evidence is not consistent with the Publishers’ claims that IA’s
distribution of the Works substituted for Plaintiffs’ sales. (See, e.g., Jorgensen Tr. at 90 -94, 96-
16
See Suppl. McN Decl. Ex. 7.
17
ECF No. 108-1 (“Jorgensen Report”) ¶38.
26
97 for summary of his limited data points.) But Jorgensen’s analysis does not show causality;
instead, it relates only to correlation. See Prince Decl. ¶¶35-53. The life cycle of every book is
unique, as innumerable factors may affect its success across formats and throughout the years,
ranging from broad macroeconomic factors and changing topics of public interest, to marketing
and pricing, to the reactions of press and readers, to school adoptions and whether school is in
session, to global interest and film adaptations. See SUMF¶¶27-29, 59-62. Jorgensen ignores all
these other factors, rendering it impossible for him to reliably separate out the impact of IA’s
NEL in Q3 2020. Prince Decl. ¶¶35-53; Jorgensen Tr. at 106-143, 159-160. Jorgensen did not
even look at whether the Works had historically varied sales by season (id. at 108:20 – 110:18) –
even though Reimers performed a seasonality check for these same Works showing that in the
prior year, Amazon print sale rankings for the Works also experienced a notable spike in Q2
(Reimers Report ¶53) – which should have sent Jorgensen back to the drawing board.
Even worse, Dr. Jorgensen failed to control for the unprecedented macroeconomic effects
occurring in the two quarters in 2020 that were his focus, including the complex impacts of
COVID on the book industry, which varied by title, and the Black Lives Matter movement and
the Presidential election season, which shifted reader interests. Prince Report ¶¶83-85, 94-99;
SUMF¶30. Jorgensen’s own graph suggests that in Q3 2020, the Publishers’ library ebook
checkouts for many of the Works were declining from the prior quarter simply because they
were reverting to pre-pandemic levels, as physical libraries and bookstores began to re-open and
lockdowns lifted – and thus that the decline in check-outs demonstrates nothing about the impact
of IA’s NEL. Prince Decl. ¶¶47-48.
27
Jorgensen’s analysis is so unreliable that even IA’s expert Reimers was forced to
acknowledge in her deposition that the sales and checkout trajectories on which Jorgensen based
his conclusions could be explained by several factors or varying differences across publishers
and that one would have to “think hard” about “how [one] can get … a causal relationship off of
the Internet Archive on … sales in general” based on “these data just by themselves.”18
3. Reimers Provides Limited and Unreliable Testimony
As for Reimers, her focus is very narrow, her conclusions are limited, and her results are
also unreliable.
18
Reimers Tr. at 267:8 - 272:20, Prince Decl. ¶39. See also Reimers Tr. at 35:10 -36:13; Reimers Report
¶35 (admitting that in light of the impact of COVID, “one cannot simply look at absolute changes in
sales”).
28
Critically, Reimers only analyzes the purported impact of IA on Amazon print sales
rankings for the Works and does not analyze ebook revenues in any market. This is a glaring
omission, given her own research concluding that ebooks are the “closest substitute” to free
infringing digital scans of print books and that print books are not “close substitutes.” Prince
Decl. ¶34.19 Further, Amazon print sales rankings are an odd focus – as Reimers admits, they are
based on an unknown algorithm, do not necessarily correlate with revenue, and do not include
major market segments including independent bookstores, wholesalers and Big Box stores, and
Reimers did nothing to confirm that they reliably reflect the broader market, while Exhibits 16a
and 16b to Jorgensen’s Report suggest they may not. Reimers Tr. at 122-124, 126-128. Finally,
although Reimers has more controls in her study than Jorgensen, her controls are still insufficient
– including because she, too, focuses primarily on 2020 but does not adequately control for the
impact of COVID or the Black Lives Matter (“BLM”) movement. Prince Decl. ¶¶54-75;
Reimers Tr. at 80-90, 179-209, 213-221.
While IA’s brief repeatedly proclaims that its expert reports “demonstrate” “that no
[market] harm has occurred” (IA Br. at 25), such hyped-up causal statements bear no relation to
Reimers’ conclusions. See Reimers Tr. at 178:4-13 (conceding that “statistical analyses aren’t
set up to definitively prove anything … including” the “study of the Amazon print sale rankings
in [her] expert report.”). Reimers conducts three analyses, which IA’s brief inaccurately
describes. First, Reimers examined IA’s impact on Amazon print rankings after the Works were
made available on IA’s website. While IA’s Brief misleadingly states as fact that “the available
19
While IA tries to blame the Publishers for Reimers’ and Jorgensen’s lack of additional financial
information, IA voluntarily withdrew its requests for such information in exchange for the Publishers’
withdrawal of other requests. Suppl. McN Decl. Ex. 8.
29
data shows that people don’t buy fewer copies of a book when that book is available for
borrowing via CDL” (IA Br. at 24), Reimers had far less certitude: her range of likely results
included a harmful impact on the Works’ Amazon print rankings when posted on IA’s Website.
Reimers Report ¶46; Reimers Tr. at 167-169; Prince Decl. ¶¶57-58. Moreover, Foster
Declaration Exhibit 6 shows that approximately half of the Works were first made available on
IA before 2018, before it significantly scaled up. Further, this says nothing about the impact of
IA if CDL becomes ubiquitous in the future. Prince Decl. ¶61.
Next, Reimers looked at the time period after the Works were removed from the NEL,
and found a decline of 1-2% in the Works’ Amazon print rankings – which, of course, could
have been caused by many factors. Prince Decl. ¶¶68-75. While IA’s Brief tries to make this
sound determinative, Reimers admits that this evidence is “weak” (Reimers Report ¶38) and that
“I cannot rule … out entirely” that IA hurt the Publishers’ sales (Reimers Tr. at 177-5-178:13).
Finally, Reimers looked at Amazon print sale ranking for the Works after the launch of
the NEL. Although she found an improvement in ranking, she had to toss out these results after
finding a similar improvement in ranking in March 2019, suggesting that “seasonality” was the
cause. Reimers Report ¶53; Prince Decl. ¶¶43, 66-67 & n.5.
In sum, all four factors of the fair use inquiry reinforce that Internet Archive has no fair
use defense.
VI. INTERNET ARCHIVE IS NOT ENTITLED TO REMITTANCE OF

STATUTORY DAMAGES UNDER SECTION 504
As a last resort, in a tacit concession of liability, IA argues that it is entitled to remittance
of statutory damages. IA Br. at 35-37. Section 504 of the Copyright Act directs courts to remit
statutory damages where the infringer is an employee or agent of a library who “infringed by
reproducing the work in copies,” and “believed and had reasonable grounds for believing” that
30
its use of the work was a fair use. 17 U.S.C. §504(c)(2). While IA’s Section 504 argument is
premature on this liability motion, it fails for several independent reasons.
First, §504’s remittance provision applies only to employees of libraries who infringe a
work “by reproducing the work in copies.” 17 U.S.C. §504(c)(2). It therefore only remits
statutory damages for infringement of the reproduction right, and offers no cover for IA’s
infringement of the distribution, public display or public performance rights.
Second, the exemption is for physical libraries and does not extend to IA. When
enacting the DMCA, Congress addressed what type of institution qualifies as a §108 library and
“ma[de] clear that … the term ‘libraries’ and ‘archives’ as used and described in this provision
still refer to such institutions only in the conventional sense of entities that are established as, and
conduct their operations through, physical premises.” S. Rep. No. 105-190, 62 (1998). It
pointedly declined to extend the exemptions “to online digital ‘libraries’ … that exist only in the
virtual (rather than physical) sense,” citing concern that “[t]he ease with which such sites are
established online literally allows anyone to create his or her own digital ‘library’” Id. In short,
Congress foresaw the danger of expanding library exemptions to entities like IA, and determined
that such unregulated institutions should not receive special dispensation based merely on their
self-identification as libraries in the digital realm.
Third, IA cannot rely on its alleged belief in the lawfulness of its conduct because it has
asserted privilege to block discovery on this very question. See, e.g., Suppl. McN Decl. Ex. 9
(“IA Priv. Log”), passim (asserting privilege over numerous communications, including to and
from Michelle Wu, concerning “attorney advice on CDL legal strategy” and “attorney advice on
NEL legal strategy”). “[A] party may not assert that it believed its conduct was lawful, and
simultaneously claim privilege to block inquiry into the basis for the party’s state of mind.”
31
Arista Records v. Lime Grp., 2011 WL 1642434, at *3 (S.D.N.Y. Apr. 20, 2011) (citing United
States v. Bilzerian, 926 F.2d 1285, 1292 (2d Cir. 1991)). Accordingly, a party who claims
attorney-client privilege over communications about the legality of its conduct is barred from
defending itself based on a belief that the conduct was lawful – even where, as here, the party
does not explicitly invoke attorney advice as the basis for its belief. As the court explained in
Arista, “it would be unfair for a party asserting contentions of good faith to then rely on its
privileges to deprive its adversary of access to material that might disprove or undermine the
party’s contentions.” Id. The logic of Arista is directly applicable here. IA deprived the
Publishers of the ability to assess the content of the advice IA received about the legality of its
unprecedented actions making it impossible for Publishers or this Court to assess whether IA had
“reasonable grounds” for believing its activities constituted fair use for the purposes of §504.
Fourth, IA’s argument that it had a good-faith belief that CDL was fair use rests entirely
on Brewster Kahle’s reliance on the White Paper and/or the Position Statement on CDL. Yet by
the time these documents were published in September 2018, IA had been distributing in-
copyright books for years (SUMF¶¶226-35, 351, 624-35) and had been operating its Open
Libraries project for several months. ECF No. 96-7 (McN Decl. Ex. 7, “Freeland Tr.”) at 106:1-
4; see also Suppl. McN Decl. Ex. 10 (“Bailey Tr.”) at 215:2-15. Moreover, IA had been notified
repeatedly over the years by the Publishers and others that its activities infringed their
copyrights. SUMF¶¶624-35. And IA ignored the advice of its own counsel for the ReDigi
amicus brief that the Second Circuit’s decision raised serious questions about the legality of
CDL. SUMF¶¶457-60. Moreover, the record emphatically demonstrates that IA did not follow
the practices prescribed in the White Paper. See supra at 11-13. If anything, the White Paper
and Statement provide strong evidence that IA did not “reasonably believe” its CDL or NEL
32
practices were permissible under copyright law since IA deviated in crucial respects from the
framework that ostensibly justified them. In the end, IA’s supposed “good faith” is the ultimate
fact question and, if IA is not precluded from relying on §504, the record is replete with contrary
evidence.
CONCLUSION
Plaintiffs respectfully request that the Court deny Internet Archive’s motion for summary
judgment, and grant Plaintiffs’ motion for summary judgment with respect to the causes of
action set forth in the Complaint.
Dated: New York, New York DAVIS WRIGHT TREMAINE LLP

September 2, 2022
/s/ Elizabeth A. McNamara
Linda Steinman
John M. Browning
Jesse Feitel
Carl Mazurek
1251 Avenue of the Americas, 21st Floor
New York, NY 10020
Phone: (212) 489-8230
Email: lizmcnamara@dwt.com
lindasteinman@dwt.com
jackbrowning@dwt.com
jessefeitel@dwt.com
carlmazurek@dwt.com
OPPENHEIM + ZEBRAK, LLP
Matthew J. Oppenheim
Scott A. Zebrak (pro hac vice)
Danae Tinelli (pro hac vice)
4530 Wisconsin Avenue, NW, 5th Floor
Washington, DC 20016
Phone: (202) 480-2999
Email: matt@oandzlaw.com
scott@oandzlaw.com
danae@oandzlaw.com
33
Attorneys for Plaintiffs

Hachette Book Group, Inc., HarperCollins
Publishers LLC, John Wiley & Sons, Inc.,
and Penguin Random House LLC
34
CERTIFICATION OF COMPLIANCE
As counsel of record to Plaintiffs Hachette Book Group, Inc., HarperCollins Publishers
LLC, John Wiley & Sons, Inc. and Penguin Random House LLC, I hereby certify that this brief
complies with the type-volume limitations set forth in this Court’s Individual Rule II.D and
complies with all formatting requirements set forth therein. I am relying upon the word count
function of the word-processing system (Microsoft Word), which indicates that 9,990 words
appear in the brief, except for the portions excluded from the word count by Rule II.D.
/s/ Elizabeth A. McNamara
35

Pltfs Brief in Opp To Defs MSJ REDACTED

Uploaded by

Copyright:

Available Formats

Pltfs Brief in Opp To Defs MSJ REDACTED

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Pltfs Brief in Opp To Defs MSJ REDACTED

Uploaded by

Copyright:

Available Formats

Case 1:20-cv-04160-JGK-OTW Document 169 Filed 09/02/22 Page 1 of 41

UNITED STATES DISTRICT COURT

PLAINTIFFS’ MEMORANDUM OF LAW IN OPPOSITION TO INTERNET

DAVIS WRIGHT TREMAINE LLP

OPPENHEIM + ZEBRAK, LLP

PRELIMINARY STATEMENT .................................................................................................... 1

I. CONTROLLED DIGITAL LENDING IS NOT A DEFENSE TO

A. Congress and the Copyright Office Have Repeatedly Considered

B. The Second Circuit and Other Courts Rejected the Premises of

1. The Owned-To-Loaned Ratio Does Not Save IA ........................... 4

2. ReDigi Rejected Internet Archive’s Argument that Fair Use

C. The Facts Disprove Internet Archive’s Presentation of CDL as a

1. Internet Archive Grossly Inflates the Prevalence of CDL .............. 9

2. Internet Archive Does Not Obey CDL’s Core Principles ............. 11

II. FACTOR ONE WEIGHS IN FAVOR OF PLAINTIFFS .................................... 13

A. IA’s Infringement Is Not Transformative ................................................. 13

B. IA Ignores the Commercial Aspects of Its Website and Operations ........ 15

III. FACTOR TWO WEIGHS IN FAVOR OF PLAINTIFFS ................................... 16

IV. FACTOR THREE WEIGHS IN FAVOR OF PLAINTIFFS ............................... 17

V. FACTOR FOUR WEIGHS IN FAVOR OF PLAINTIFFS ................................. 18

A. Internet Archive Fails to Pay Fees that Libraries Customarily Pay

B. Undisputed Evidence Establishes that IA Offers A Competing

1. IA’s Experts Do Not Rebut the Evidence of Market Harm .......... 25

2. Jorgensen’s Analysis is Unreliable ............................................... 26

3. Reimers Provides Limited and Unreliable Testimony .................. 28

VI. INTERNET ARCHIVE IS NOT ENTITLED TO REMITTANCE OF

CERTIFICATION OF COMPLIANCE ....................................................................................... 35

A&M Records v. Napster,

Arista Records v. Lime Grp.,

Association of Am. Publishers, Inc. v. Frosh,

Authors Guild, Inc. v. HathiTrust,

Authors Guild v. Google,

Bureau of Nat’l Literature v. Sells,

Campbell v. Acuff-Rose Music,

Capitol Records, LLC v. ReDigi Inc.,

City of Providence v. Bats Global Markets,

Disney Enterprises v. VidAngel,

Doan v. American Book Co.,

Fox News Network v. TVEyes, Inc.,

Harper & Row Publishers v. Nation Enters.,

In re Electronic Books Antitrust Litigation,

Infinity Broad. Corp. v. Kirkwood,

Lamarr-Arruz v. CVS Pharmacy,

On Davis v. The Gap,

Oracle Corp. v. SAP AG,

Penguin Grp. (USA) v. Am. Buddha,

Ringgold v. Black Entm’t Television,

Soc’y of Holy Transfiguration Monastery v. Gregory,

Sony Corp. of Am. v. Universal City Studios,

Twin Peaks Prods., Inc. v. Publ’ns Int’l,

UMG Recordings v. MP3.com, Inc.,

Worldwide Church of God v. Phila. Church of God,

F.R.E. 702 ......................................................................................................................................18

144 Cong. Rec. H.7074 (1998) ........................................................................................................3

S. Rep. 105-190 (1998) ..................................................................................................................31

like Sandra Cisneros, who rely on copyright to secure their livelihoods.

I. CONTROLLED DIGITAL LENDING IS NOT A DEFENSE TO COPYRIGHT

reproduction, of millions of unauthorized ebooks. To the contrary, §109(a) states that

permits its practice of CDL.