Memorandum in Opposition re [214] MOTION to Compel and for In Camera Review REDACTED VERSION filed by Google LLC. (Attachments: # (1) Exhibit A SEALED Declaration, # (2) Exhibit B SEALED, # (3) Exhibit C SEALED)(Reilly, Craig)
Page 1 PageID#
UNITED STATES DISTRICT COURT
EASTERN DISTRICT OF VIRGINIA
ALEXANDRIA DIVISION
UNITED STATES, et al.,
Plaintiffs,
vs.
No. 1:23-cv-00108-LMB-JFA
GOOGLE LLC,
Defendant.
GOOGLE LLC’S MEMORANDUM OF LAW IN OPPOSITION
TO PLAINTIFFS’ MOTION FOR IN CAMERA INSPECTION AND
TO COMPEL PRODUCTION OF DOCUMENTS WITHHELD AS PRIVILEGED
Plaintiffs’ Motion fails to satisfy the standard required for in camera inspection in this
District. Where, as here, Google has provided Plaintiffs both a detailed privilege log and
30(b)(6) deposition testimony that establishes the factual basis for each of its claimed assertions
of privilege, Plaintiffs do not have a right to an in camera review. Plaintiffs can obtain in
camera review only if they present “‘a factual basis sufficient to support a reasonable, good
faith belief that in camera inspection may reveal evidence that information in the materials is
not privileged.’” In re Zetia (Ezetimibe) Antitrust Litig., 2019 WL 6122012, at *4 (E.D. Va.
July 16, 2019). Nor should this Court reward Plaintiffs’ delay in challenging Google’s
privilege and work product claims of which they have been aware for almost two years,
particularly given that the United States had the opportunity over a year ago to ask about the
basis for Google’s privilege claims regarding the Remedy Projects when they deposed a Google
30(b)(6) witness for a full day on this very topic. Plaintiffs’ assertion that the Court should
now spend its time to inspect in camera an initial sample set of 21 documents lacks support in
REDACTED VERSIONPage 2 PageID#
the record and fails because Plaintiffs have not met their burden to come forward with a
sufficient factual basis to challenge Google’s privilege assertions.
First, 17 of the 21 documents at issue are either fully protected from disclosure or
properly contain redactions based on, at a minimum, the work-product doctrine, because they
relate to or reveal the substantive details of projects Google undertook in direct response to
then-active government investigations and with a view to possibly resolving those
investigations or, if not, to resolving anticipated litigation.1 Each of those projects involved
specific efforts to explore a range of potential remedies that could resolve the active
investigations and avoid litigation. This is classic work product.
Second, two of the remaining four documents are protected from disclosure by the
attorney-client privilege because they involve either requests for or the provision of legal
advice.
Plaintiffs’ primary basis for questioning Google’s claims of privilege is their
allegation, imported from news reports and allegations in other cases, that the documents are
possibly being withheld on the basis of a “Communicate with Care training” that Google
employees received.3 Such speculation does not constitute a factual basis sufficient to support
Google has in good faith re-reviewed these 17 documents and determined that it could narrow
its assertions of privilege over an identical redaction in two of the documents. Google has
reproduced these two documents to Plaintiffs.
Google has in good faith re-reviewed these four documents and determined that it could
narrow its assertions of privilege over two of the documents and produce the other two in full.
Google has reproduced these four documents to Plaintiffs.
The United States made a similar “Communicate with Care” argument in its search case
against Google, and Judge Mehta not only denied the United States’ motion to compel and for
sanctions; he mentioned Google’s “good-faith effort” to consider any challenges to entries on
its log. See Transcript of Status Conf. at 62, United States v. Google LLC, 1:20-cv-03010-APM
(D.D.C. May 12, 2022), ECF No. 353 (rejecting DOJ’s request for mass production of
challenged documents because “there was no way I was going to . . . turn over those kind of
records en masse to an adversary, given what seems to us to have been a good-faith effort by
Google to go back and look at these records and produce them, re-review them, re-categorize
them, de-privilege those that need to be de-privileged, and produce them to the plaintiffs”).
And Plaintiffs’ reference to sanctions in In re Google Play Store Antitrust Litigation is a red
herring: although the opinion mentioned “Communicate with Care,” 2023 WL 2673109, at
REDACTED VERSIONPage 3 PageID#
in camera inspection. Whether or not the document contained phrases such as “privileged and
confidential” or “communicate with care,” Google reviewed each and every document to make
a separate privilege determination. An exhibit Plaintiffs cite in their Motion illustrates this
very point:
Google produced the document, despite the fact that it had been labeled
“privileged” by the participants in the communication, because of the independent privilege
review Google undertook. See Mot. Ex. 5.
Finally, Plaintiffs’ request for an in camera inspection of 21 documents that they claim
are only a “sample” and, based on the Court’s review, may warrant further in camera
inspections, has no basis in the law. It is tantamount to an improper request for an advisory
opinion. Privilege is determined on a document-by-document basis. Yet Plaintiffs seek to
bypass the required document-by-document determination by presenting to the Court a cherrypicked collection of documents. The Court should decline to make any broad pronouncements
about Google’s approach to privilege and instead decide only the privilege claims relating to
the particular documents that are the subject of the Motion. As demonstrated below, the
documents still at issue are protected by the attorney-client privilege or the work-product
doctrine (or both), and Plaintiffs’ Motion should be denied.
BACKGROUND
Google’s ad tech business has been the subject of regulatory investigations since early
2019.4 Several of the investigations have since ripened into litigation, including this case, while
*2-3 (N.D. Cal. Mar. 28, 2023), the sanctioned conduct there involved an entirely separate
issue, id. at *7-10 (preservation of chat communications)—one that is not before this Court.
These include investigations by the U.K. Information Commissioner’s Office (initiated
February 2019); the Irish Data Protection Commission (initiated May 2019); the French
Competition Authority (initiated July 2019); the U.K. Competition and Markets Authority
(initiated July 2019); the State of Texas, leading a coalition of state attorneys general (initiated
September 2019); the Antitrust Division of the United States Department of Justice (initiated
October 2019); and the European Commission (initiated November 2019). See Opp. Ex. A,
Declaration of Theodore Lazarus ¶ 5 (“Lazarus Decl.”); Mot. Ex. 29, Ltr. from J. Elmer, Feb.
REDACTED VERSIONPage 4 PageID#
others have concluded with settlements based on remedies packages offered by Google and
accepted by the regulator.5 Complying with numerous requests for information and Civil
Investigative Demands (“CIDs”) on the aggressive timetables demanded by multiple
government authorities was a huge undertaking, with Google producing nearly three million
documents to the DOJ Antitrust Division alone.
The present dispute arises from Google’s production of documents pursuant to multiple
CIDs issued by the Antitrust Division during its lengthy investigation. Google devoted
immense resources to reviewing for privilege when producing documents in response to the
CIDs. The process involved first- and second-level reviews, re-reviews, and various levels of
quality assurance, the latter of which resulted in post-production clawbacks and throwbacks,
as is not uncommon in document productions of this size. Google’s outside counsel and
contract attorneys have spent more than 42,000 hours in privilege-related work, and Google
has updated and supplemented its privilege logs as it continued to produce more documents.
The resulting privilege log identified roughly 180,000 documents as privileged—
approximately 6% of Google’s total production. Of the nearly three million documents
produced to DOJ, Google clawed back 4,900. See Mot. 23. Google’s privilege reviews were
hardly a rote, box-checking exercise. On the contrary, Google produced many documents
labeled “privileged and confidential” and even communications involving attorneys, based on
25, 2022 at 3; see also Opp. Ex. B, Google 30(b)(6) Dep. 71:19-20 (“Google’s position is
consistent with everything that's contained in this Feb. 25 letter.”).
See, e.g., The Autorité de la Concurrence Hands Out a €220 Millions Fine to Google for
Favouring Its Own Services in the Online Advertising Sector, Autorité de la Concurrence (June
7, 2021), https://shorturl.at/oquL2.
REDACTED VERSIONPage 5 PageID#
its document-by-document review in light of applicable law, as Plaintiffs themselves
acknowledge, see Mot. 12.In response to numerous active government investigations of its ad tech business and
anticipated litigation, Google considered potential remedies in order to resolve those
investigations and avoid litigation. Beginning in late 2019, Google initiated a number of
projects to develop and evaluate those potential remedies: Projects SingleClick, Stonehenge,
Sunday, Monday, and Banksy (the “Remedy Projects”). Plaintiff United States has known
about these Remedy Projects, and their asserted privileged nature, at least since August 19,
2021, when outside counsel for Google sent a letter to DOJ identifying materials created as
part of these projects as being protected by the attorney-client privilege and work-product
doctrine. See Mot. Ex. 11, Ltr. from J. Elmer, Aug. 19, 2021. That letter began a lengthy
period of correspondence, lasting through July 2022, between Google and DOJ regarding
Google’s privilege claims over the Remedy Projects. During this period, on February 28, 2022,
DOJ deposed a Google corporate representative regarding the company’s privilege and workproduct assertions with respect to the Remedy Projects. See Opp. Ex. B, 30(b)(6) Dep. of
Alphabet at 61:8-62:7, 69:16-71:22 (“Google 30(b)(6) Dep.”); see also Mot. Ex. 29, Ltr. from
J. Elmer, Feb. 25, 2022; Opp. Ex. B, Google 30(b)(6) Dep. 71:19-20 (“Google’s position is
consistent with everything that’s contained in this Feb. 25 letter. And I am also here to ask
[sic] any additional questions that you may have.”).
As Google’s corporate representative testified, each Remedy Project was an analysis of
potential remedies undertaken because of and in response to active government investigations
by regulators and in anticipation of litigation concerning Google’s ad tech business. See Opp.
In fact, when it came to Google’s attention that some of the redactions in six of the documents
at issue here were overbroad, Google promptly provided Plaintiffs with complete or less
redacted versions of the affected documents. See supra nn.1-2.
REDACTED VERSIONPage 6 PageID#
Ex. B, Google 30(b)(6) Dep. 109:2-6, 153:7-11 (Project SingleClick); id. at 149:13-25, 188:24189:4 (Project Stonehenge); id. at 72:17-22, 94:6-12 (Project Sunday); id. at 121:6-13, 132:1024 (Project Monday); id. at 201:5-7, 201:17-204:20, 211:5-6 (Project Banksy); Opp. Ex. A,
Lazarus Decl. ¶¶ 5-9. Accordingly, the Remedy Projects and the documents associated with
them are classic work product.
Following Google’s 30(b)(6) deposition, the United States took no steps—such as the
filing of a CID enforcement action, as was its prerogative under 15 U.S.C. § 1314—to
challenge Google’s assertion of privilege, and largely stopped corresponding with Google over
these claims for nearly a year.7 Yet only now—nearly two years after having been notified of
Google’s privilege claims over these documents—have Plaintiffs moved to compel production
of these documents.
ARGUMENT
The Court should see Plaintiffs’ Motion for what it is: an improper attempt to use a
narrow set of cherry-picked documents as a key to unlock a much broader swathe of documents
reflecting Google’s strategic legal thinking about how to respond to active regulatory
investigations (and potential litigation) regarding its ad tech business—including remedies that
Google considered to possibly settle those investigations and anticipated litigation. This is
classic work product, to which Plaintiffs are not entitled. Plaintiffs’ Motion for in camera
inspection lacks merit for the reasons set forth below. The Court should deny the Motion.
I.
Plaintiffs Fail to Justify In Camera Review
Google—through its detailed privilege logs and correspondence, as well as its 30(b)(6)
deposition—has carried its initial burden of demonstrating that the attorney-client privilege
DOJ tried to revive its long-abandoned challenge to Google’s privilege and work-product
assertions by sending a letter to counsel for Google a week before filing suit, see Mot. Ex. 1,
Ltr. from B. Nakamura, Jan. 18, 2023, but then filed suit less than a week after doing so, without
giving Google any meaningful chance to respond.
REDACTED VERSIONPage 7 PageID#
and/or work product doctrine apply to the documents at issue. See United States v. Jones, F.2d 1069, 1072 (4th Cir. 1982). Because Google has carried that burden, Plaintiffs now must
come forth with evidence to rebut Google’s showing. Zetia, 2019 WL 6122012, at *(“Plaintiffs must make a threshold ‘showing of a factual basis adequate to support a good faith
belief by a reasonable person’ that an exception to the privilege may apply.”); accord NLRB v.
Interbake Foods, LLC, 637 F.3d 492, 502 (4th Cir. 2011) (“Once a prima facie showing of a
privilege has been made, an opposing party can justify in camera inspection of the documents
by advancing ‘a factual basis sufficient to support a reasonable, good faith belief that in camera
inspection may reveal evidence that information in the materials is not privileged.’” (citation
omitted)); Allstate Ins. Co. v. Warns, 2013 WL 1310556, at *2 (D. Md. Mar. 28, 2013) (same);
Brown Univ. v. Tharpe, 2012 WL 12894480, at *3 (E.D. Va. Mar. 30, 2012) (same).
For three reasons, Plaintiffs lack an adequate factual basis:
First, Google has established that it initiated the Remedy Projects in anticipation of
litigation stemming from multiple active and ongoing regulatory investigations—a
straightforward application of the work-product doctrine. In addition to its privilege logs,
Google has supported its privilege and work-product claims regarding the Remedy Projects
through other extrinsic evidence, including a 30(b)(6) deposition and a declaration submitted
with this Opposition. Plaintiffs try to sidestep the issue by arguing that because some of the
Remedy Projects bore a “nexus” to “strategic business decisions,” the Remedy Projects were
“business projects,” whether or not they were initiated and conducted in anticipation of
litigation. Mot. 15. But as discussed below, infra pp. 9, 11-13, that is not the law.
Second, Plaintiffs’ general discussion of Google’s “Communicate with Care” training
fails to support their contention that the documents at issue in their Motion are not privileged.
Mot. 14-15 (inviting the Court to review the documents “through the lens of Google’s
Communicate with Care policy”).
Google’s privilege determinations are based on the
REDACTED VERSIONPage 8 PageID#
substance of each document as part of its document-by-document review, and it is not claiming
privilege on the basis of any “Communicate with Care” policy or training. Google’s production
of documents labeled “privileged” proves as much and illustrates the flimsiness of Plaintiffs’
argument. Plaintiffs’ speculation that documents may not be privileged does not justify in
camera review. In re NC Swine Farm Nuisance Litig., 2017 WL 2313470, at *6 (E.D.N.C.
May 26, 2017) (“[M]ere speculation that the purpose was business and not legal does not justify
in camera review.”).
Third, the fact that Google has clawed back documents, as permitted by both the Federal
Rules and the governing Protective Order in this case, does not warrant in camera review. See
Fed. R. Evid. 502(b); Modified Prot. Order ¶ 12, ECF No. 203. Given the size and speed of
the document productions it was required to make during the DOJ’s investigation, Google’s
production process and its clawbacks have been manifestly reasonable. Courts in the Fourth
Circuit and across the country recognize that, in electronic document reviews of this size and
complexity—ranging in the millions of documents—perfection is impossible and privileged
material will slip through a reasonable review. See, e.g., Zetia, 2019 WL 6122012, at *(reasoning that “Rule 502 was intended to permit . . . relative efficiencies” in privilege review
“while preserving privilege protections in the event of isolated disclosure”); In re Grand Jury
Investigation, 142 F.R.D. 276, 280 (M.D.N.C. 1992) (recognizing that “perfect precaution” is
not “required” in conducting privilege reviews (emphasis omitted)). In fact, clawing back only
4,900 documents out of a production volume of roughly 2.9 million (a clawback rate of 0.17%)
reflects a robust and thorough privilege review. See Zetia, 2019 WL 6122012, at *9 (clawback
rate of 0.003% reflected “an entirely reasonable error rate” compared to an unreasonable rate
of approximately 30%). Google has also reasonably and promptly pursued its clawbacks. For
example, in its clawback efforts over the Remedy Projects at issue in this Motion, once
Google’s outside counsel learned of the privileged nature of the projects in August 2021,
REDACTED VERSIONPage 9 PageID#
Google promptly sent letters to DOJ on a rolling basis to identify and claw back the
inadvertently disclosed documents relating to those projects.Plaintiffs’ mere suspicion or, quite frankly, distrust of Google’s work product and
privilege determinations does not provide a basis to burden this Court with in camera review.
Glaxo, Inc. v. Novopharm Ltd., 148 F.R.D. 535, 540 n.3 (E.D.N.C. 1993) (“If this court were
to review each and every document withheld as privileged in litigation . . . for no reason other
than counsel’s distrust of his adversary, this courthouse could hardly function.”).
II.
Documents Relating to the Remedy Projects Are Work Product.
The work-product doctrine protects documents or tangible things prepared (i) in
anticipation of litigation or for trial (ii) by or for a party (i.e., by or for a party or a party’s
representative). Fed. R. Civ. P. 26(b)(3). Documents created as part of the Remedy Projects
or that describe or reveal the substance of the Remedy Projects satisfy both elements.
See, e.g., Mot. Ex. 11, Ltr. from J. Elmer, Aug. 19, 2021. Although Plaintiffs speculate that
Google has clawed back documents only after realizing their “evidentiary value,” Mot. 23, the
facts say otherwise. Google has engaged in multiple privilege re-reviews, which has led it not
only to claw back 4,900 documents but also to throw back (i.e., de-privilege or lift redactions
for) over 79,000 more. Google does not deny that some of its 4,900 clawbacks were instigated
after Google’s counsel scrutinized the document when Plaintiffs sought to introduce it at a
deposition. There is nothing improper with clawing back a document after becoming aware of
its privileged nature when the opposing party places the document at issue. In fact, Rule 502(b)
exists for this precise reason: to allow parties to remedy “inadvertent” disclosures without
waiving their claims of privilege. Fed. R. Civ. P. 502(b). Any other holding would effectively
require producing parties to meticulously re-review every single produced document for fear
of waiver if the opposing party brings it up first. And furthermore, this scenario accounts for
a small minority of the clawbacks that Google has requested.
Google is defending 17 of the 21 challenged documents on the basis of work-product
protection. These 17 include Appendix Documents 1 (GOOG-DOJ-12766025), 2 (GOOGDOJ-15231660), 5 (GOOG-DOJ-AT-01106088), 6 (GOOG-DOJ-AT-01106174), 8 (GOOGDOJ-AT-01498627),
(GOOG-DOJ-AT-01514569),
(GOOG-DOJ-AT01692081), 11 (GOOG-DOJ-AT-01887780), 12 (GOOG-DOJ-AT-01895607), 13 (GOOGDOJ-AT-01914586), 14 (GOOG-DOJ-AT-02099758), 15 (GOOG-DOJ-AT-00660900), (GOOG-DOJ-AT-00205841), 17 (GOOG-DOJ-AT-00030150), 18 (GOOG-DOJ-AT00660895), 19 (GOOG-DOJ-AT-01007207), and 20 (GOOG-DOJ-AT-01687296). Google
has in good faith removed an identical redaction in Documents 15 (GOOG-DOJ-AT00660900) and 18 (GOOG-DOJ-AT-00660895), because the material related to the non
REDACTED VERSIONPage 10 PageID#
In anticipation of litigation. The Remedy Project documents meet the first element,
which asks whether the document was “prepared because of the prospect of litigation when the
preparer faces an actual claim or a potential claim following an actual event or series of events
that reasonably could result in litigation.” Nat’l Union Fire Ins. Co. of Pittsburgh, Pa. v.
Murray Sheet Metal Co., 967 F.2d 980, 984 (4th Cir. 1992); accord, e.g., Sher v. Barclays Cap.
Inc., 2013 WL 3279801, at *3 (D. Md. June 26, 2013).10 This element, at its core, distinguishes
“materials prepared in the ordinary course of business or pursuant to regulatory requirements
or for another non-litigation purpose” from those prepared “in anticipation of litigation.” Nat’l
Union, 967 F.2d at 984.
As Google’s corporate representative testified, each of the five challenged Remedy
Projects was initiated because of active and ongoing government investigations that could ripen
into litigation, satisfying the anticipation-of-litigation requirement. See Opp. Ex. B, Google
30(b)(6) Dep. 109:2-5, 144:13-145:5, 146:11-13, 151:5-17 (Project SingleClick was initiated
by a Google in-house attorney and a Google manager in December 2019, supervised by Google
in-house counsel, and “would never have been undertaken if [Google] did not anticipate
litigation”); id. at 160:6-18, 188:20-23, 190:22-191:2 (Project Stonehenge was undertaken “to
privileged “Banksy” as opposed to the work-product-protected Project Banksy, see infra n.16,
and has provided revised versions of these documents to Plaintiffs. See supra n.1.
Courts construe “litigation” broadly and include active regulatory investigations as equating
to anticipated litigation. For instance, a district court in this Circuit has found material to be
work product when the material was created because of anticipated regulatory investigations.
See Adair v. EQT Prod. Co., 294 F.R.D. 1, 5-6 (W.D. Va. 2013). Courts outside of this Circuit
are in accord. See, e.g., United States ex rel. Baker v. Cmty. Health Sys., Inc., 2012 WL
12342816, at *3 (D.N.M. May 16, 2012) (“[O]nce a government investigation is begun, many
courts concede that litigation may not be far behind. And therefore documents that are prepared
after a government investigation is already underway may well be accorded work-product
protection”); In re Grand Jury Subpoena, 220 F.R.D. 130, 147 (D. Mass. 2004) (finding “once
a governmental investigation has begun, litigation is sufficiently likely to satisfy the
‘anticipation’ requirement”); Galvin v. Hoblock, 2003 WL 22208370, at *3-4 (S.D.N.Y. Sept.
24, 2003) (“[T]he term ‘litigation’ encompasses not only litigation in court, but also quasijudicial proceedings before a government agency.”).
REDACTED VERSIONPage 11 PageID#
be prepared in case of any litigation” resulting from identified regulatory investigations, and
supervised by Google in-house counsel and involved U.S. and European outside counsel); id.
at 73:7-11, 76:13-14, 88:14-15, 89:11-12, 96:16-98:3 (Project Sunday was initiated by a
Google-in house attorney,11 involved Google employees working “as specifically directed by
legal,” and undertaken to “start mapping out analyses” of remedy options in order “[t]o
understand the implications of [Google’s] potential responses to regulatory actions” that
Google was “very concerned” would ripen into litigation); id. at 121:6-13, 123:3-14, 124:1518 (Project Monday “was an analysis for a particular remedy to be undertaken due to . . .
anticipated litigation” and included a small number of businesspeople working “[i]n
conjunction with legal counsel [to] author[] a potential remedy”); id. at 200:22-201:7, 204:916, 205:6-17, 209:1-19, 211:5-6 (Project Banksy was undertaken because of the French
Competition Authority’s investigation as “a potential settlement with a regulatory authority,”
was part of remedies package French Competition Authority ultimately accepted, and was led
and supervised by Google in-house counsel). Because “the driving force behind the preparation
of” the Remedy Project documents was anticipated litigation, the work-product doctrine shields
them from disclosure. Nat’l Union, 967 F.2d at 984.
Plaintiffs’ reliance on the individual deposition of
in claiming that
lawyers were not involved in Project Sunday, see Mot. 20-21 (citing Ex. 30, Dep. of
at 243:4-248:25), is misplaced for several reasons. First,
himself
testified that Ted Lazarus, an in-house Google attorney, was one of the members in charge of
the “day-to-day” direction of Project Sunday. See Mot. Ex. 30, Dep. of
at 249:1-5. Second, DOJ requested, and Google put forward, a 30(b)(6) corporate
representative specifically to give testimony on the Remedy Projects at issue, including Project
Sunday. That corporate representative testified that Ted Lazarus, the same in-house attorney
mentioned, initiated Project Sunday, Opp. Ex. B, Google 30(b)(6) Dep. 88:1415, and that
’s role in Project Sunday was in fact fairly limited, see id. at 85:1286:13. While the testimony of
and Google’s corporate representative are not
inconsistent, any marginal discrepancy between the two should be read in light of, and defer
to, the 30(b)(6) testimony as representing Google’s official position, which is also corroborated
by the Declaration of Mr. Lazarus himself. See Opp. Ex. A, Lazarus Decl. ¶ 7.
REDACTED VERSIONPage 12 PageID#
Courts in this District and elsewhere routinely reach the same conclusion in analogous
contexts involving materials generated as part of a settlement analysis. In Zetia, for example,
the court held that settlement analyses performed by financial analysts at a company embroiled
in litigation was entitled to work-product protection. 2019 WL 6122012, at *3-4. It did not
matter that the documents “reflect[ed] the kind of business analysis the company would
perform irrespective of pending litigation.” Id. at *2, *4. The work-product doctrine still
applied, since the company undertook the analyses only to evaluate the effect of settling a
litigation. Id. at *3-4 (“[H]ad counsel ‘not requested this analysis in order to evaluate a
potential settlement, the [document] and the analysis within it would not have been created.’”).
Similarly, work-product materials do not lose their work-product status because they
incorporate ordinary-course analyses or content.
See id. at *5 (finding financial data
incorporated into work-product document was entitled to the same work product protection
afforded to the entire document because the data was “integral to the privileged purpose of the
document” even when the data, standing alone, would not have otherwise been privileged);
Jud. Watch, Inc. v. DOJ, 432 F.3d 366, 371 (D.C. Cir. 2005) (“[F]actual material is itself
privileged when it appears within documents that are attorney work product.”).12 Accordingly,
the fact that a preexisting analysis by an investment banking firm (over which Google is not
claiming work-product protection) was incorporated as a factual input into Project Sunday
materials, see Opp. Ex. B, Google 30(b)(6) Dep. 116:12-117:6, does not undermine the workproduct nature of the Project Sunday materials. See Mot. 16, 20.
See also, e,g., Aerojet Rocketdyne, Inc. v. Glob. Aerospace, Inc., 2019 WL 4929930, at *(E.D. Cal. Oct. 7, 2019) (denying in camera review where it was “arguable that the [document]
as a whole was entitled to protection even if every fact or number it contained was not
independently privileged.”); CFPB v. Ocwen Fin. Corp., 2019 WL 1119788, at *3 (S.D. Fla.
Mar. 12, 2019) (preexisting ordinary-course Powerpoint slides were privileged where “the
specific factual information was curated and reformulated by attorneys” and reflected legal
strategy).
REDACTED VERSIONPage 13 PageID#
Plaintiffs’ argument that the Remedy Projects are not work product because they bear
a “clear nexus to ordinary-course strategic business decisions,” Mot. 15, misstates the
applicable legal standard. The anticipation-of-litigation element hinges on why a document
was created, see Nat’l Union, 967 F.2d at 984, not on whether it has a “nexus” to business
decisions. Under the “because of” test—which applies in this Circuit, id., as well as the D.C.
and Second Circuits—“material generated in anticipation of litigation may also be used for
ordinary business purposes without losing its protected status.” United States v. Deloitte LLP,
610 F.3d 129, 138 (D.C. Cir. 2010); see also United States v. Adlman, 134 F.3d 1194, 1195 (2d
Cir. 1998) (“[A] document . . . does not lose work-product protection merely because it is
intended to assist in the making of a business decision influenced by the likely outcome of the
anticipated litigation.”). As a result, courts in circuits that apply the “because of” test
consistently hold that settlement-evaluation materials are work product, even if they also
inform or lead to business decision-making.In the related MDL proceeding pending in the Southern District of New York, Judge
Castel has already deemed references to the substance of one of the Remedy Projects—Project
SingleClick—as protected work product.
He credited a declaration from a Google
See FTC v Boehringer (Boehringer I), 778 F.3d 142, 150 (D.C. Cir. 2015) (“We find no
merit in the proposition that any settlement term that has some independent economic value to
both parties must always be treated as an ordinary (non-litigation) business transaction for
purposes of work product protection. Common sense and practical experience teach that
settlement deals routinely include arrangements that could be isolated from the overall
agreement and stand on their own but were nonetheless crafted for the purpose of settling
litigation.”); FTC v Boehringer (Boehringer II), 892 F.3d 1264, 1268 (D.C. Cir. 2018) (holding
that business analysis prepared to assist counsel in evaluating a settlement was work product
even though evaluating the settlement “ultimately was a business decision as well as a legal
decision”); In re Payment Card Interchange Fee & Merch. Disc. Antitrust Litig., 2018 WL
1162552, at *6-11 (E.D.N.Y. Feb. 26, 2018) (holding that settlement workstream documents
“created to assess business implications of the litigation or settlement” and drafted by nonlawyer at the direction of counsel were opinion work product); Bovis Lend Lease, LMB, Inc. v.
Seasons Contracting Corp., 2002 WL 31729693, at *6 (S.D.N.Y. Dec. 5, 2002) (finding
litigation strategy document was still protected by work product doctrine even where it referred
to potential business decisions resulting from the litigation).
REDACTED VERSIONPage 14 PageID#
businessperson that Project SingleClick was an analysis of actions undertaken in response to
active regulatory investigations, and involved the “subject of liability avoidance” which
reflected the “‘mental impressions, conclusions, opinions, or legal theories of an attorney’
prepared ‘in anticipation of litigation.’” In re Google Digit. Advert. Antitrust Litig., 2023 WL
196146, at *3 (S.D.N.Y. Jan. 17, 2023) (quoting Fed. R. Civ. P. 26(b)(3)). Judge Castel did
not have access to the Google 30(b)(6) deposition testimony regarding the Remedy Projects;
however, the Google declaration that he reviewed explained the project just as the 30(b)(6)
deposition did. Compare id. (citing the declaration as explaining that Project SingleClick was
“an analysis of certain actions undertaken ‘in response to active investigations into, and in
anticipation of litigation concerning, Google’s AdTech business’”), with Opp. Ex. B, Google
30(b)(6) Dep. 144:19-24 (describing Project SingleClick as “an analysis of potential remedies
to some anticipated regulatory actions . . . undertaken in anticipation of litigation concerning
Google’s ad tech business”). And while Judge Castel did not rule that every document referring
to the work of Project SingleClick was categorically protected—as he properly limited his
ruling to the specific documents before him—his reasoning that the work-product doctrine
protects projects undertaken because of anticipated litigation and reflecting legal settlement
strategy applies with equal force here.
By or for a party. The Remedy Project documents also meet the work-product
doctrine’s second element, which asks whether the document was prepared “by or for [the
withholding] party or its representative (including the … party’s attorney, consultant, surety,
indemnitor, insurer, or agent).” Fed. R. Civ. P. 26(b)(3). Plaintiffs cannot dispute that the
Remedy Project documents were created by or for Google.
Instead, Plaintiffs attempt to graft an additional element onto the work-product
doctrine: that the document be prepared “at the direction of an attorney.” Mot. 8 (quoting ePlus
Inc. v. Lawson Software, Inc., 280 F.R.D. 247, 251 (E.D. Va. 2012)). But nowhere does Rule
REDACTED VERSIONPage 15 PageID#
26(b)(3) require that the document be prepared by or at the direction of an attorney. See Hempel
v. Cydan Dev., Inc., 2020 WL 4933634, at *4 (D. Md. Aug. 24, 2020) (“[T]he statutory text of
Rule 26 does not require that documents be ‘prepared by, for, or at the request of [a] lawyer’
in order to be protected as work product.”). The work-product doctrine, as the Fourth Circuit
has recognized, protects documents prepared in anticipation of litigation “whether the material
was actually prepared by [an] attorney or by another ‘representative’ of the [withholding]
party.” Nat’l Union, 967 F.2d at 984.As explained in the accompanying Lazarus Declaration, each of the Remedy Project
documents was either prepared in response to active regulatory investigations or reveals
substantive details of the Remedy Projects. See Opp. Ex. A, Lazarus Decl. ¶¶ 5-9. And as
explained in that Declaration, Google’s in-house counsel initiated, oversaw, and/or played a
key role in each of the Remedy Projects. See id.; see also Opp. Ex. B, Google 30(b)(6) Dep.
74:19-75:5, 88:14-15, 122:16-123:18, 128:11-129:9, 146:11-13, 149:7-12, 151:5-17, 182:16-
As this Court and others have recognized, work product extends to materials beyond those
prepared by, or at the direction of, attorneys. See Dunston v. Cecil Huang, 2010 WL 11698098,
at *1 (E.D. Va. May 5, 2010) (Anderson, M.J.) (holding that material “created by a
representative of the defendants” in anticipation of litigation was “protected by the work
product protection”); see also, e.g., Williams v. AT&T Mobility, 2022 WL 2821922, at *(E.D.N.C. July 19, 2022) (rejecting argument that “since the documents were created by nonlawyers, they cannot be privileged [as work product]”); Coffey v. Tyler Staffing Servs., Inc.,
2020 WL 1307233, at *3 (W.D. Va. Mar. 19, 2020) (“The fact that these materials were not
prepared by an attorney does not, by itself, disqualify them from falling under the work product
doctrine.”); Gutshall v. New Prime, Inc., 196 F.R.D. 43, 46 (W.D. Va. 2000) (finding materials
that party itself “prepared, or commissioned the preparation of” constituted work product);
Haigh v. Matsushita Elec. Corp. of Am., 676 F. Supp. 1332, 1356-58 (E.D. Va. 1987) (rejecting
defendants’ argument that tape recordings were not work product because they were made
absent attorney direction, explaining that “Defendants are living in the past, and are presenting
a pre-1970 argument”—i.e., from before Rule 26(b)(3) was amended). To the extent that ePlus
suggests otherwise, Mot. 8, it contradicts the plain language of Rule 26(b)(3). Relying solely
on Hickman v. Taylor, 329 U.S. 495 (1947), see ePlus, 280 F.R.D. at 257, the ePlus court failed
to consider that the amendments to Rule 26(b)(3) expanded the work-product doctrine. Moore
v. Tri-City Hosp. Auth., 118 F.R.D. 646, 649 (N.D. Ga. 1988) (“Rule 26(b)(3) . . . preserves
the essential portions of the doctrine as announced in Hickman, but also notably expands the
doctrine by extending discovery protection to the work product of a party as well as that party’s
attorneys.”).
REDACTED VERSIONPage 16 PageID#
183:1, which further corroborates that the Remedy Projects were undertaken in anticipation of
litigation and reflect legal strategy. This is analogous to Zetia, where the court found potential
settlement analyses to be work product in part because they were initiated by and overseen by
in-house counsel. See 2019 WL 6122012, at *3.
Plaintiffs mischaracterize Google’s work-product assertions over Remedy Project
documents as shielding “major business projects simply because they were undertaken while
the company was under government investigation.” Mot. 7. But Google has consistently made
clear, through extensive correspondence with DOJ,15 the February 28, 2022 deposition of its
corporate representative, and a meet-and-confer with DOJ in May of that year, that the basis
for its assertion of work-product protection for the Remedy Project documents is that they were
created because of active government investigations and anticipated litigation.16 This is exactly
what the law requires for the work-product protection to apply. Each of the 17 Remedy Project
documents was either created as part of or reveals the substance of at least one of the Remedy
See, e.g., Mot. Ex. 29, Ltr. from J. Elmer, Feb. 25, 2022.
Although one of the Remedy Projects—Project Banksy—involved the analysis, for the
purpose of potentially settling a regulatory investigation, of a concept sourced in a nonprivileged, prior workstream referred to simply as “Banksy,” that does not preclude Project
Banksy from receiving work-product protection. Project Banksy involved the analysis and
eventual execution of some of the ideas from non-privileged “Banksy” in the specific context
of the French Competition Authority and other regulators’ investigations. See Opp. Ex. C,
Dep. of
at 168:1-172:10 (describing “Banksy”); Opp. Ex. B, Google 30(b)(6)
Dep. 200:22-201:7 (explaining Project Banksy as emerging from “Banksy” in response to the
French Competition Authority’s investigation). That analysis and execution may not have been
undertaken at all, and certainly would not have been undertaken in substantially similar form,
but for the French Competition Authority investigation and Google’s legal strategy to identify
settlement options in response to that investigation. See Opp. Ex. B, Google 30(b)(6) Dep.
210:15-211:6. Moreover, Google has asserted privilege over only Project Banksy materials,
but not prior “Banksy” materials, which underscores that Google has properly claimed workproduct protection over only those documents reflecting the company’s legal strategy and
where the active regulatory investigations and anticipated litigation were the driving force
behind the documents’ creation.
REDACTED VERSIONPage 17 PageID#
Projects, in whole or in part, see Opp. Ex. A, Lazarus Decl. ¶¶ 11-15, and has therefore been
properly redacted or withheld.
III.
Google Has Properly Redacted or Withheld Attorney-Client Privileged
Documents
The remaining two documents at issue concern the attorney-client privilege.17 The
singular basis for Plaintiffs’ challenge to these documents is Google’s “Communicate with
Care” training. According to Plaintiffs, this training caused some Google employees to label
communications privileged and confidential and include an attorney even when the
communication was not privileged. Mot. 11-12. From there, Plaintiffs speculate that some of
the documents Google has withheld in this case must have been the product of “Communicate
with Care,” rather than of good-faith privilege determinations. There is no basis for such
speculation. As Google has explained to Plaintiffs, Google has not claimed privilege based on
“Communicate with Care,” nor has Google withheld documents on the basis of privilege
simply because they contain the words “Privileged and Confidential.” Google has reviewed
documents one by one and has made good faith privilege determinations based on the substance
of each document even within a voluminous and fast paced production. A document Plaintiffs
appended as an exhibit to their Motion, see Mot. Ex. 5, underscores Google’s focus on a
document’s substance—and not whether it is labeled privileged. See Mot. 12 (observing
produced document appended as Exhibit 5 contained a privilege legend, but lacked “any
request for legal advice”).
Plaintiffs’ challenges to these two documents do not reflect any pervasive issues with
Google’s privilege log or Google’s application of the attorney-client privilege:
These two documents—excluding the ones that Google has thrown back, see supra n.2—are
Documents 3 (GOOG-DOJ-AT-00029680) and 7 (GOOG-DOJ-AT-01139556).
REDACTED VERSIONPage 18 PageID#
Document 3 (GOOG-DOJ-AT-00029680). Google has in good faith lifted the second
redaction at -685 and the redaction at -686. The document now contains a single redaction at
-685, containing feedback on the deal terms from “Legal,” which “contain[s] legal advice of
ESQ regarding contract terms,” as noted in the privilege log.
Document 7 (GOOG-DOJ-AT-01139556). This is a 47-page presentation titled
“Streamline programmatic buying doors on AdMob and Ad Manager,” containing redactions
at -564, -569, 570, -579, -589, and -602.18 The redacted portions on -564, -579, and -602 are
comments seeking, containing, and responding to legal advice given by Google’s in-house
counsel
about information on the previous slide (-563, and duplicated at -
and -601), detailing revenue share scenarios for AdMob and Ad Manager. The redaction at
-569 contains “[f]eedback from legal”—plainly facially privileged material—and the redacted
rows on -570 convey “[r]egulatory [r]isk” and potential mitigating actions reflecting legal
advice of
.
In addition, while Plaintiffs take issue with the privilege-log
description for these redactions, they do not explain why that description—“seeking and
revealing legal advice of
ESQ regarding legal aspects of product
development”—is “materially deficient.” Mot. 24.Regarding the redaction at -589, after re-reviewing this document in good faith, Google
has removed this redaction of a single slide within a 47-page presentation and has provided
Plaintiffs with a replacement copy of this document, see supra n.2.
The three pages -564, -579, and -602 are exact copies of each other and the redactions were
identical except for a portion at the bottom of -564, which had been inadvertently left
unredacted. Google has taken steps to conform the redactions on -564 to the redactions on -and -602 and has reproduced the document to Plaintiffs accordingly.
Plaintiffs claim that no lawyer appears in the metadata for this document. That claim is
incorrect, as there are eight attorneys in the presentation’s metadata, including
and outside counsel.
REDACTED VERSIONPage 19 PageID#
Documents 4 (GOOG-DOJ-AT-01019463) and 21 (GOOG-DOJ-06588656). Google
has reviewed these two documents and has in good faith opted to drop its claims of privilege
over these two documents. As with the voluntarily lifted targeted redactions in Documents 3,
7, 15, and 18, lifting the redactions from Documents 4 and 21 reflects Google’s consistent and
ongoing good faith and reasonable efforts to balance the production of relevant material with
zealously protecting material that is legally privileged. See Transcript of Status Conf. at 62,
United States v. Google LLC, No. 1:20-cv-03010-APM (D.D.C. Oct. 20, 2020), ECF No. (viewing Google’s throwbacks as part of a “good-faith effort to . . . re-review [withheld
documents], re-categorize them, de-privilege those that need to be de-privileged, and produce
them to the plaintiffs”); Google Digit. Advert., 2023 WL 196146, at *3 (Castel, J.) (“The Court
declines to ascribe any nefarious motive to [Google’s] change of position” by no longer
asserting privilege over a challenged document); see also Am. Nat’l Bank & Tr. Co. of Chi. v.
Equitable Life Assurance Soc’y of U.S., 406 F.3d 867, 870, 879 (7th Cir. 2005) (defendant
“exhibited good faith throughout the privilege log proceedings [by] among other acts,” rereviewing and throwing back documents as necessary).
IV.
Plaintiffs Improperly Seek an Advisory Opinion
The Court should decline to grant Plaintiffs the advisory opinion they seek. Although
Plaintiffs contend that they have selected a “representative” sample of documents, Mot. 1, they
have cherry-picked 21 documents out of nearly 180,000 documents withheld or redacted over
the course of a three-year investigation, each dissimilar and covering a range of issues, from
the millions of documents produced by Google. From this small and cherry-picked sample, it
is not possible to extrapolate any conclusions about Google’s overall production and privilege
log.
Plaintiffs’ request that the Court treat a review of 21 documents as “exemplars of three
categories of documents,” Mot. 4, is not how privilege determinations work. Rather, privilege
REDACTED VERSIONPage 20 PageID#
review involves “a document-by-document privilege analysis.” Interbake, 637 F.3d at (“[E]ach e-mail within a particular line of discussion must be analyzed separately for privilege
purposes.”); see also In re Rivastigmine Pat. Litig., 2005 WL 2319005, at *5 (S.D.N.Y. Sept.
22, 2005) (declining to rule on privilege not tied to “specific documents” because “privilege
determination is fact-sensitive”). In the parallel MDL proceeding, Judge Castel rejected a
similar request by those plaintiffs to “clarify” how the attorney-client privilege would apply to
documents not submitted for in camera review, concluding that it was “essentially a request
for an advisory opinion on legal principles.” Google Digit. Advert., 2023 WL 196146, at *1.
The same reasoning applies equally here.
V.
Google’s Exercise of Its Clawback Rights Does Not Justify In Camera Review
Plaintiffs try to use the fact that Google clawed back documents to argue that the Court
should review the 21 documents in camera. But the fact that Google has exercised its right to
claw back inadvertently produced privileged material has no bearing on whether in camera
review is appropriate. See supra Point I. Plaintiffs cite no case law suggesting otherwise.VI.
Google Has Not Waived Privilege over Document 1 (GOOG-DOJ-12766025).
Plaintiffs contend that Google has waived privilege over one document (GOOG-DOJ-
12766025),21 by failing to object or claw back material read into the record during a Google
Plaintiffs cannot use the fact that Google has clawed back documents as a backdoor to in
camera review. Plaintiffs say that because “all of the documents identified in the Appendix
have been clawed back at one time or another,” in camera review will help them “articulate
fully their challenge to Google’s claims of privilege.” Mot. 24. Not only is that not the standard
for in camera review, see supra Point I, but it also makes little sense. It is, of course, true
across all privilege challenges—whether privilege was asserted along with an initial production
or asserted by clawback—that the opposing party has a limited ability to understand the
withheld content and make arguments against privilege. But if the challenging party could
force in camera review simply by claiming that it needed to see materials withheld as privileged
in order to challenge even more withheld materials, then the in camera standard would be
meaningless.
In Document 1 (GOOG-DOJ-12766025), Google redacted communications revealing the
substance of Projects SingleClick and Stonehenge. Because Projects SingleClick and
Stonehenge refer to Remedy Projects initiated because of anticipated litigation, and producing
REDACTED VERSIONPage 21 PageID#
employee’s August 10, 2021 deposition. See Mot. 19 n.33. But Plaintiffs mischaracterize the
record. Google did not allow portions of the document to be read into the record at the
deposition without objection. Instead, DOJ read a portion of the document into the record as
part of a question, and before counsel could instruct the witness not to answer and claw back
the document, the witness responded. See Mot. Ex. 18, Dep. of
at
197:17-23. When DOJ sought to clarify the witness’s answer, counsel for Google objected.
Id. at 197:24-198:8. After a short break, DOJ restated its question, and counsel for Google
instructed the witness not to answer on the basis of privilege. Id. at 198:16-199:3. The
document read into the record was then immediately clawed back as privileged during the
deposition. Id. at 199:4-11, 201:22-202:2.
These facts make it clear that Google did not waive privilege. On the contrary,
Google’s counsel immediately informed DOJ during the deposition that the exhibit contained
privileged material with respect to Projects Stonehenge and SingleClick, and promptly took
steps to rectify the disclosure by clawing back the document and instructing the witness not to
answer additional questions. These were reasonable steps to protect privilege in light of an
inadvertent production. See, e.g., King Pharm., Inc. v. Purdue Pharma, L.P., 2010 WL
2243872, at *2 (W.D. Va. June 2, 2010) (“Purdue clearly acted promptly in taking reasonable
steps to rectify the error” when it clawed back, during a deposition, “one of millions of
documents produced by the parties in [a] complex [] case.”).
Moreover, the Google
employee, in his individual capacity, could not waive privilege or work-product protection on
the document unredacted would reveal Google settlement strategy and analysis, the material is
properly redacted as work product. See supra Point II.
See also Karma Auto. LLC v. Lordstown Motors Corp., 2021 WL 4147007, at *3 (C.D. Cal.
June 16, 2021) (holding that privilege was not waived where counsel “object[ed] on the record
at [witness’s] deposition and request[ed] that [non-disclosing party] destroy and refrain from
further use of the document”).
REDACTED VERSIONPage 22 PageID#
behalf of the company by answering a question before the company’s counsel had time to
object. “[T]he issue of waiver in the case of the testimony of a corporation’s agent is whether
the corporation has acted deliberately in disclosing privileged information, not whether the
witness acted deliberately.” Parneros v. Barnes & Noble, Inc., 332 F.R.D. 482, 504-(S.D.N.Y. 2019) (emphases added). And nothing in the record suggests that Google intended
to waive privilege. Google has consistently taken the position that documents relating to the
Remedy Projects are protected work product.
CONCLUSION
For the foregoing reasons, the Court should deny Plaintiffs’ Motion.
Dated: June 2,
Respectfully submitted,
Eric Mahr (pro hac vice)
Andrew Ewalt (pro hac vice)
Julie Elmer (pro hac vice)
Lauren Kaplin (pro hac vice)
Jeanette Bayoumi (pro hac vice)
Claire Leonard (pro hac vice)
Sara Salem (pro hac vice)
Tyler Garrett (VSB # 94759)
FRESHFIELDS BRUCKHAUS
DERINGER US LLP
700 13th Street, NW, 10th Floor
Washington, DC Telephone: (202) 777-Facsimile: (202) 777-eric.mahr@freshfields.com
/s/ Craig C. Reilly
Craig C. Reilly (VSB # 20942)
THE LAW OFFICE OF
CRAIG C. REILLY, ESQ.
209 Madison Street
Alexandria, VA Telephone: (703) 549-Facsimile: (703) 549-craig.reilly@ccreillylaw.com
Daniel Bitton (pro hac vice)
AXINN, VELTROP & HARKRIDER LLP
55 2nd Street
San Francisco, CA Telephone: (415) 490-Facsimile: (415) 490-dbitton@axinn.com
Bradley Justus (VSB # 80533)
Koren Wong-Ervin (pro hac vice)
AXINN, VELTROP & HARKRIDER LLP
1901 L Street, NW
Karen L. Dunn (pro hac vice)
Jeannie H. Rhee (pro hac vice)
William A. Isaacson (pro hac vice)
Joseph Bial (pro hac vice)
Amy J. Mauser (pro hac vice)
Martha L. Goodman (pro hac vice)
Bryon P. Becker (VSB #93384)
Erica Spevack (pro hac vice)
PAUL, WEISS, RIFKIND, WHARTON &
GARRISON LLP
2001 K Street, NW
Washington, DC 20006-Telephone: (202) 223-Facsimile (202) 223-kdunn@paulweiss.com
Meredith Dearborn (pro hac vice)
REDACTED VERSIONPage 23 PageID#
Washington, DC Telephone: (202) 912-Facsimile: (202) 912-bjustus@axinn.com
PAUL, WEISS, RIFKIND, WHARTON &
GARRISON LLP
535 Mission Street, 24th Floor
San Francisco, CA Telephone: (646) 432-Facsimile: (202) 330-mdearnborn@paulweiss.com
Counsel for Defendant Google LLC
REDACTED VERSION
PDF Page 1
PlainSite Cover Page
PDF Page 2
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 1 of 23 PageID# 2048
UNITED STATES DISTRICT COURT
EASTERN DISTRICT OF VIRGINIA
ALEXANDRIA DIVISION
UNITED STATES, et al.,
Plaintiffs,
vs.
No. 1:23-cv-00108-LMB-JFA
GOOGLE LLC,
Defendant.
GOOGLE LLC’S MEMORANDUM OF LAW IN OPPOSITION
TO PLAINTIFFS’ MOTION FOR IN CAMERA INSPECTION AND
TO COMPEL PRODUCTION OF DOCUMENTS WITHHELD AS PRIVILEGED
Plaintiffs’ Motion fails to satisfy the standard required for in camera inspection in this
District. Where, as here, Google has provided Plaintiffs both a detailed privilege log and
30(b)(6) deposition testimony that establishes the factual basis for each of its claimed assertions
of privilege, Plaintiffs do not have a right to an in camera review. Plaintiffs can obtain in
camera review only if they present “‘a factual basis sufficient to support a reasonable, good
faith belief that in camera inspection may reveal evidence that information in the materials is
not privileged.’” In re Zetia (Ezetimibe) Antitrust Litig., 2019 WL 6122012, at *4 (E.D. Va.
July 16, 2019). Nor should this Court reward Plaintiffs’ delay in challenging Google’s
privilege and work product claims of which they have been aware for almost two years,
particularly given that the United States had the opportunity over a year ago to ask about the
basis for Google’s privilege claims regarding the Remedy Projects when they deposed a Google
30(b)(6) witness for a full day on this very topic. Plaintiffs’ assertion that the Court should
now spend its time to inspect in camera an initial sample set of 21 documents lacks support in
1
REDACTED VERSION
PDF Page 3
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 2 of 23 PageID# 2049
the record and fails because Plaintiffs have not met their burden to come forward with a
sufficient factual basis to challenge Google’s privilege assertions.
First, 17 of the 21 documents at issue are either fully protected from disclosure or
properly contain redactions based on, at a minimum, the work-product doctrine, because they
relate to or reveal the substantive details of projects Google undertook in direct response to
then-active government investigations and with a view to possibly resolving those
investigations or, if not, to resolving anticipated litigation.1 Each of those projects involved
specific efforts to explore a range of potential remedies that could resolve the active
investigations and avoid litigation. This is classic work product.
Second, two of the remaining four documents are protected from disclosure by the
attorney-client privilege because they involve either requests for or the provision of legal
advice.2
Plaintiffs’ primary basis for questioning Google’s claims of privilege is their
allegation, imported from news reports and allegations in other cases, that the documents are
possibly being withheld on the basis of a “Communicate with Care training” that Google
employees received.3 Such speculation does not constitute a factual basis sufficient to support
1
Google has in good faith re-reviewed these 17 documents and determined that it could narrow
its assertions of privilege over an identical redaction in two of the documents. Google has
reproduced these two documents to Plaintiffs.
2
Google has in good faith re-reviewed these four documents and determined that it could
narrow its assertions of privilege over two of the documents and produce the other two in full.
Google has reproduced these four documents to Plaintiffs.
3
The United States made a similar “Communicate with Care” argument in its search case
against Google, and Judge Mehta not only denied the United States’ motion to compel and for
sanctions; he mentioned Google’s “good-faith effort” to consider any challenges to entries on
its log. See Transcript of Status Conf. at 62, United States v. Google LLC, 1:20-cv-03010-APM
(D.D.C. May 12, 2022), ECF No. 353 (rejecting DOJ’s request for mass production of
challenged documents because “there was no way I was going to . . . turn over those kind of
records en masse to an adversary, given what seems to us to have been a good-faith effort by
Google to go back and look at these records and produce them, re-review them, re-categorize
them, de-privilege those that need to be de-privileged, and produce them to the plaintiffs”).
And Plaintiffs’ reference to sanctions in In re Google Play Store Antitrust Litigation is a red
herring: although the opinion mentioned “Communicate with Care,” 2023 WL 2673109, at
2
REDACTED VERSION
PDF Page 4
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 3 of 23 PageID# 2050
in camera inspection. Whether or not the document contained phrases such as “privileged and
confidential” or “communicate with care,” Google reviewed each and every document to make
a separate privilege determination. An exhibit Plaintiffs cite in their Motion illustrates this
very point:
Google produced the document, despite the fact that it had been labeled
“privileged” by the participants in the communication, because of the independent privilege
review Google undertook. See Mot. Ex. 5.
Finally, Plaintiffs’ request for an in camera inspection of 21 documents that they claim
are only a “sample” and, based on the Court’s review, may warrant further in camera
inspections, has no basis in the law. It is tantamount to an improper request for an advisory
opinion. Privilege is determined on a document-by-document basis. Yet Plaintiffs seek to
bypass the required document-by-document determination by presenting to the Court a cherrypicked collection of documents. The Court should decline to make any broad pronouncements
about Google’s approach to privilege and instead decide only the privilege claims relating to
the particular documents that are the subject of the Motion. As demonstrated below, the
documents still at issue are protected by the attorney-client privilege or the work-product
doctrine (or both), and Plaintiffs’ Motion should be denied.
BACKGROUND
Google’s ad tech business has been the subject of regulatory investigations since early
2019.4 Several of the investigations have since ripened into litigation, including this case, while
*2-3 (N.D. Cal. Mar. 28, 2023), the sanctioned conduct there involved an entirely separate
issue, id. at *7-10 (preservation of chat communications)—one that is not before this Court.
4
These include investigations by the U.K. Information Commissioner’s Office (initiated
February 2019); the Irish Data Protection Commission (initiated May 2019); the French
Competition Authority (initiated July 2019); the U.K. Competition and Markets Authority
(initiated July 2019); the State of Texas, leading a coalition of state attorneys general (initiated
September 2019); the Antitrust Division of the United States Department of Justice (initiated
October 2019); and the European Commission (initiated November 2019). See Opp. Ex. A,
Declaration of Theodore Lazarus ¶ 5 (“Lazarus Decl.”); Mot. Ex. 29, Ltr. from J. Elmer, Feb.
3
REDACTED VERSION
PDF Page 5
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 4 of 23 PageID# 2051
others have concluded with settlements based on remedies packages offered by Google and
accepted by the regulator.5 Complying with numerous requests for information and Civil
Investigative Demands (“CIDs”) on the aggressive timetables demanded by multiple
government authorities was a huge undertaking, with Google producing nearly three million
documents to the DOJ Antitrust Division alone.
The present dispute arises from Google’s production of documents pursuant to multiple
CIDs issued by the Antitrust Division during its lengthy investigation. Google devoted
immense resources to reviewing for privilege when producing documents in response to the
CIDs. The process involved first- and second-level reviews, re-reviews, and various levels of
quality assurance, the latter of which resulted in post-production clawbacks and throwbacks,
as is not uncommon in document productions of this size. Google’s outside counsel and
contract attorneys have spent more than 42,000 hours in privilege-related work, and Google
has updated and supplemented its privilege logs as it continued to produce more documents.
The resulting privilege log identified roughly 180,000 documents as privileged—
approximately 6% of Google’s total production. Of the nearly three million documents
produced to DOJ, Google clawed back 4,900. See Mot. 23. Google’s privilege reviews were
hardly a rote, box-checking exercise. On the contrary, Google produced many documents
labeled “privileged and confidential” and even communications involving attorneys, based on
25, 2022 at 3; see also Opp. Ex. B, Google 30(b)(6) Dep. 71:19-20 (“Google’s position is
consistent with everything that's contained in this Feb. 25 letter.”).
See, e.g., The Autorité de la Concurrence Hands Out a €220 Millions Fine to Google for
Favouring Its Own Services in the Online Advertising Sector, Autorité de la Concurrence (June
7, 2021), https://shorturl.at/oquL2.
5
4
REDACTED VERSION
PDF Page 6
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 5 of 23 PageID# 2052
its document-by-document review in light of applicable law, as Plaintiffs themselves
acknowledge, see Mot. 12.6
In response to numerous active government investigations of its ad tech business and
anticipated litigation, Google considered potential remedies in order to resolve those
investigations and avoid litigation. Beginning in late 2019, Google initiated a number of
projects to develop and evaluate those potential remedies: Projects SingleClick, Stonehenge,
Sunday, Monday, and Banksy (the “Remedy Projects”). Plaintiff United States has known
about these Remedy Projects, and their asserted privileged nature, at least since August 19,
2021, when outside counsel for Google sent a letter to DOJ identifying materials created as
part of these projects as being protected by the attorney-client privilege and work-product
doctrine. See Mot. Ex. 11, Ltr. from J. Elmer, Aug. 19, 2021. That letter began a lengthy
period of correspondence, lasting through July 2022, between Google and DOJ regarding
Google’s privilege claims over the Remedy Projects. During this period, on February 28, 2022,
DOJ deposed a Google corporate representative regarding the company’s privilege and workproduct assertions with respect to the Remedy Projects. See Opp. Ex. B, 30(b)(6) Dep. of
Alphabet at 61:8-62:7, 69:16-71:22 (“Google 30(b)(6) Dep.”); see also Mot. Ex. 29, Ltr. from
J. Elmer, Feb. 25, 2022; Opp. Ex. B, Google 30(b)(6) Dep. 71:19-20 (“Google’s position is
consistent with everything that’s contained in this Feb. 25 letter. And I am also here to ask
[sic] any additional questions that you may have.”).
As Google’s corporate representative testified, each Remedy Project was an analysis of
potential remedies undertaken because of and in response to active government investigations
by regulators and in anticipation of litigation concerning Google’s ad tech business. See Opp.
6
In fact, when it came to Google’s attention that some of the redactions in six of the documents
at issue here were overbroad, Google promptly provided Plaintiffs with complete or less
redacted versions of the affected documents. See supra nn.1-2.
5
REDACTED VERSION
PDF Page 7
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 6 of 23 PageID# 2053
Ex. B, Google 30(b)(6) Dep. 109:2-6, 153:7-11 (Project SingleClick); id. at 149:13-25, 188:24189:4 (Project Stonehenge); id. at 72:17-22, 94:6-12 (Project Sunday); id. at 121:6-13, 132:1024 (Project Monday); id. at 201:5-7, 201:17-204:20, 211:5-6 (Project Banksy); Opp. Ex. A,
Lazarus Decl. ¶¶ 5-9. Accordingly, the Remedy Projects and the documents associated with
them are classic work product.
Following Google’s 30(b)(6) deposition, the United States took no steps—such as the
filing of a CID enforcement action, as was its prerogative under 15 U.S.C. § 1314—to
challenge Google’s assertion of privilege, and largely stopped corresponding with Google over
these claims for nearly a year.7 Yet only now—nearly two years after having been notified of
Google’s privilege claims over these documents—have Plaintiffs moved to compel production
of these documents.
ARGUMENT
The Court should see Plaintiffs’ Motion for what it is: an improper attempt to use a
narrow set of cherry-picked documents as a key to unlock a much broader swathe of documents
reflecting Google’s strategic legal thinking about how to respond to active regulatory
investigations (and potential litigation) regarding its ad tech business—including remedies that
Google considered to possibly settle those investigations and anticipated litigation. This is
classic work product, to which Plaintiffs are not entitled. Plaintiffs’ Motion for in camera
inspection lacks merit for the reasons set forth below. The Court should deny the Motion.
I.
Plaintiffs Fail to Justify In Camera Review
Google—through its detailed privilege logs and correspondence, as well as its 30(b)(6)
deposition—has carried its initial burden of demonstrating that the attorney-client privilege
7
DOJ tried to revive its long-abandoned challenge to Google’s privilege and work-product
assertions by sending a letter to counsel for Google a week before filing suit, see Mot. Ex. 1,
Ltr. from B. Nakamura, Jan. 18, 2023, but then filed suit less than a week after doing so, without
giving Google any meaningful chance to respond.
6
REDACTED VERSION
PDF Page 8
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 7 of 23 PageID# 2054
and/or work product doctrine apply to the documents at issue. See United States v. Jones, 696
F.2d 1069, 1072 (4th Cir. 1982). Because Google has carried that burden, Plaintiffs now must
come forth with evidence to rebut Google’s showing. Zetia, 2019 WL 6122012, at *4
(“Plaintiffs must make a threshold ‘showing of a factual basis adequate to support a good faith
belief by a reasonable person’ that an exception to the privilege may apply.”); accord NLRB v.
Interbake Foods, LLC, 637 F.3d 492, 502 (4th Cir. 2011) (“Once a prima facie showing of a
privilege has been made, an opposing party can justify in camera inspection of the documents
by advancing ‘a factual basis sufficient to support a reasonable, good faith belief that in camera
inspection may reveal evidence that information in the materials is not privileged.’” (citation
omitted)); Allstate Ins. Co. v. Warns, 2013 WL 1310556, at *2 (D. Md. Mar. 28, 2013) (same);
Brown Univ. v. Tharpe, 2012 WL 12894480, at *3 (E.D. Va. Mar. 30, 2012) (same).
For three reasons, Plaintiffs lack an adequate factual basis:
First, Google has established that it initiated the Remedy Projects in anticipation of
litigation stemming from multiple active and ongoing regulatory investigations—a
straightforward application of the work-product doctrine. In addition to its privilege logs,
Google has supported its privilege and work-product claims regarding the Remedy Projects
through other extrinsic evidence, including a 30(b)(6) deposition and a declaration submitted
with this Opposition. Plaintiffs try to sidestep the issue by arguing that because some of the
Remedy Projects bore a “nexus” to “strategic business decisions,” the Remedy Projects were
“business projects,” whether or not they were initiated and conducted in anticipation of
litigation. Mot. 15. But as discussed below, infra pp. 9, 11-13, that is not the law.
Second, Plaintiffs’ general discussion of Google’s “Communicate with Care” training
fails to support their contention that the documents at issue in their Motion are not privileged.
Mot. 14-15 (inviting the Court to review the documents “through the lens of Google’s
Communicate with Care policy”).
Google’s privilege determinations are based on the
7
REDACTED VERSION
PDF Page 9
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 8 of 23 PageID# 2055
substance of each document as part of its document-by-document review, and it is not claiming
privilege on the basis of any “Communicate with Care” policy or training. Google’s production
of documents labeled “privileged” proves as much and illustrates the flimsiness of Plaintiffs’
argument. Plaintiffs’ speculation that documents may not be privileged does not justify in
camera review. In re NC Swine Farm Nuisance Litig., 2017 WL 2313470, at *6 (E.D.N.C.
May 26, 2017) (“[M]ere speculation that the purpose was business and not legal does not justify
in camera review.”).
Third, the fact that Google has clawed back documents, as permitted by both the Federal
Rules and the governing Protective Order in this case, does not warrant in camera review. See
Fed. R. Evid. 502(b); Modified Prot. Order ¶ 12, ECF No. 203. Given the size and speed of
the document productions it was required to make during the DOJ’s investigation, Google’s
production process and its clawbacks have been manifestly reasonable. Courts in the Fourth
Circuit and across the country recognize that, in electronic document reviews of this size and
complexity—ranging in the millions of documents—perfection is impossible and privileged
material will slip through a reasonable review. See, e.g., Zetia, 2019 WL 6122012, at *10
(reasoning that “Rule 502 was intended to permit . . . relative efficiencies” in privilege review
“while preserving privilege protections in the event of isolated disclosure”); In re Grand Jury
Investigation, 142 F.R.D. 276, 280 (M.D.N.C. 1992) (recognizing that “perfect precaution” is
not “required” in conducting privilege reviews (emphasis omitted)). In fact, clawing back only
4,900 documents out of a production volume of roughly 2.9 million (a clawback rate of 0.17%)
reflects a robust and thorough privilege review. See Zetia, 2019 WL 6122012, at *9 (clawback
rate of 0.003% reflected “an entirely reasonable error rate” compared to an unreasonable rate
of approximately 30%). Google has also reasonably and promptly pursued its clawbacks. For
example, in its clawback efforts over the Remedy Projects at issue in this Motion, once
Google’s outside counsel learned of the privileged nature of the projects in August 2021,
8
REDACTED VERSION
PDF Page 10
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 9 of 23 PageID# 2056
Google promptly sent letters to DOJ on a rolling basis to identify and claw back the
inadvertently disclosed documents relating to those projects.8
Plaintiffs’ mere suspicion or, quite frankly, distrust of Google’s work product and
privilege determinations does not provide a basis to burden this Court with in camera review.
Glaxo, Inc. v. Novopharm Ltd., 148 F.R.D. 535, 540 n.3 (E.D.N.C. 1993) (“If this court were
to review each and every document withheld as privileged in litigation . . . for no reason other
than counsel’s distrust of his adversary, this courthouse could hardly function.”).
II.
Documents Relating to the Remedy Projects Are Work Product.
The work-product doctrine protects documents or tangible things prepared (i) in
anticipation of litigation or for trial (ii) by or for a party (i.e., by or for a party or a party’s
representative). Fed. R. Civ. P. 26(b)(3). Documents created as part of the Remedy Projects
or that describe or reveal the substance of the Remedy Projects satisfy both elements.9
8
See, e.g., Mot. Ex. 11, Ltr. from J. Elmer, Aug. 19, 2021. Although Plaintiffs speculate that
Google has clawed back documents only after realizing their “evidentiary value,” Mot. 23, the
facts say otherwise. Google has engaged in multiple privilege re-reviews, which has led it not
only to claw back 4,900 documents but also to throw back (i.e., de-privilege or lift redactions
for) over 79,000 more. Google does not deny that some of its 4,900 clawbacks were instigated
after Google’s counsel scrutinized the document when Plaintiffs sought to introduce it at a
deposition. There is nothing improper with clawing back a document after becoming aware of
its privileged nature when the opposing party places the document at issue. In fact, Rule 502(b)
exists for this precise reason: to allow parties to remedy “inadvertent” disclosures without
waiving their claims of privilege. Fed. R. Civ. P. 502(b). Any other holding would effectively
require producing parties to meticulously re-review every single produced document for fear
of waiver if the opposing party brings it up first. And furthermore, this scenario accounts for
a small minority of the clawbacks that Google has requested.
9
Google is defending 17 of the 21 challenged documents on the basis of work-product
protection. These 17 include Appendix Documents 1 (GOOG-DOJ-12766025), 2 (GOOGDOJ-15231660), 5 (GOOG-DOJ-AT-01106088), 6 (GOOG-DOJ-AT-01106174), 8 (GOOGDOJ-AT-01498627),
9
(GOOG-DOJ-AT-01514569),
10
(GOOG-DOJ-AT01692081), 11 (GOOG-DOJ-AT-01887780), 12 (GOOG-DOJ-AT-01895607), 13 (GOOGDOJ-AT-01914586), 14 (GOOG-DOJ-AT-02099758), 15 (GOOG-DOJ-AT-00660900), 16
(GOOG-DOJ-AT-00205841), 17 (GOOG-DOJ-AT-00030150), 18 (GOOG-DOJ-AT00660895), 19 (GOOG-DOJ-AT-01007207), and 20 (GOOG-DOJ-AT-01687296). Google
has in good faith removed an identical redaction in Documents 15 (GOOG-DOJ-AT00660900) and 18 (GOOG-DOJ-AT-00660895), because the material related to the non9
REDACTED VERSION
PDF Page 11
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 10 of 23 PageID# 2057
In anticipation of litigation. The Remedy Project documents meet the first element,
which asks whether the document was “prepared because of the prospect of litigation when the
preparer faces an actual claim or a potential claim following an actual event or series of events
that reasonably could result in litigation.” Nat’l Union Fire Ins. Co. of Pittsburgh, Pa. v.
Murray Sheet Metal Co., 967 F.2d 980, 984 (4th Cir. 1992); accord, e.g., Sher v. Barclays Cap.
Inc., 2013 WL 3279801, at *3 (D. Md. June 26, 2013).10 This element, at its core, distinguishes
“materials prepared in the ordinary course of business or pursuant to regulatory requirements
or for another non-litigation purpose” from those prepared “in anticipation of litigation.” Nat’l
Union, 967 F.2d at 984.
As Google’s corporate representative testified, each of the five challenged Remedy
Projects was initiated because of active and ongoing government investigations that could ripen
into litigation, satisfying the anticipation-of-litigation requirement. See Opp. Ex. B, Google
30(b)(6) Dep. 109:2-5, 144:13-145:5, 146:11-13, 151:5-17 (Project SingleClick was initiated
by a Google in-house attorney and a Google manager in December 2019, supervised by Google
in-house counsel, and “would never have been undertaken if [Google] did not anticipate
litigation”); id. at 160:6-18, 188:20-23, 190:22-191:2 (Project Stonehenge was undertaken “to
privileged “Banksy” as opposed to the work-product-protected Project Banksy, see infra n.16,
and has provided revised versions of these documents to Plaintiffs. See supra n.1.
10
Courts construe “litigation” broadly and include active regulatory investigations as equating
to anticipated litigation. For instance, a district court in this Circuit has found material to be
work product when the material was created because of anticipated regulatory investigations.
See Adair v. EQT Prod. Co., 294 F.R.D. 1, 5-6 (W.D. Va. 2013). Courts outside of this Circuit
are in accord. See, e.g., United States ex rel. Baker v. Cmty. Health Sys., Inc., 2012 WL
12342816, at *3 (D.N.M. May 16, 2012) (“[O]nce a government investigation is begun, many
courts concede that litigation may not be far behind. And therefore documents that are prepared
after a government investigation is already underway may well be accorded work-product
protection”); In re Grand Jury Subpoena, 220 F.R.D. 130, 147 (D. Mass. 2004) (finding “once
a governmental investigation has begun, litigation is sufficiently likely to satisfy the
‘anticipation’ requirement”); Galvin v. Hoblock, 2003 WL 22208370, at *3-4 (S.D.N.Y. Sept.
24, 2003) (“[T]he term ‘litigation’ encompasses not only litigation in court, but also quasijudicial proceedings before a government agency.”).
10
REDACTED VERSION
PDF Page 12
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 11 of 23 PageID# 2058
be prepared in case of any litigation” resulting from identified regulatory investigations, and
supervised by Google in-house counsel and involved U.S. and European outside counsel); id.
at 73:7-11, 76:13-14, 88:14-15, 89:11-12, 96:16-98:3 (Project Sunday was initiated by a
Google-in house attorney,11 involved Google employees working “as specifically directed by
legal,” and undertaken to “start mapping out analyses” of remedy options in order “[t]o
understand the implications of [Google’s] potential responses to regulatory actions” that
Google was “very concerned” would ripen into litigation); id. at 121:6-13, 123:3-14, 124:1518 (Project Monday “was an analysis for a particular remedy to be undertaken due to . . .
anticipated litigation” and included a small number of businesspeople working “[i]n
conjunction with legal counsel [to] author[] a potential remedy”); id. at 200:22-201:7, 204:916, 205:6-17, 209:1-19, 211:5-6 (Project Banksy was undertaken because of the French
Competition Authority’s investigation as “a potential settlement with a regulatory authority,”
was part of remedies package French Competition Authority ultimately accepted, and was led
and supervised by Google in-house counsel). Because “the driving force behind the preparation
of” the Remedy Project documents was anticipated litigation, the work-product doctrine shields
them from disclosure. Nat’l Union, 967 F.2d at 984.
11
Plaintiffs’ reliance on the individual deposition of
in claiming that
lawyers were not involved in Project Sunday, see Mot. 20-21 (citing Ex. 30, Dep. of
at 243:4-248:25), is misplaced for several reasons. First,
himself
testified that Ted Lazarus, an in-house Google attorney, was one of the members in charge of
the “day-to-day” direction of Project Sunday. See Mot. Ex. 30, Dep. of
at 249:1-5. Second, DOJ requested, and Google put forward, a 30(b)(6) corporate
representative specifically to give testimony on the Remedy Projects at issue, including Project
Sunday. That corporate representative testified that Ted Lazarus, the same in-house attorney
mentioned, initiated Project Sunday, Opp. Ex. B, Google 30(b)(6) Dep. 88:1415, and that
’s role in Project Sunday was in fact fairly limited, see id. at 85:1286:13. While the testimony of
and Google’s corporate representative are not
inconsistent, any marginal discrepancy between the two should be read in light of, and defer
to, the 30(b)(6) testimony as representing Google’s official position, which is also corroborated
by the Declaration of Mr. Lazarus himself. See Opp. Ex. A, Lazarus Decl. ¶ 7.
11
REDACTED VERSION
PDF Page 13
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 12 of 23 PageID# 2059
Courts in this District and elsewhere routinely reach the same conclusion in analogous
contexts involving materials generated as part of a settlement analysis. In Zetia, for example,
the court held that settlement analyses performed by financial analysts at a company embroiled
in litigation was entitled to work-product protection. 2019 WL 6122012, at *3-4. It did not
matter that the documents “reflect[ed] the kind of business analysis the company would
perform irrespective of pending litigation.” Id. at *2, *4. The work-product doctrine still
applied, since the company undertook the analyses only to evaluate the effect of settling a
litigation. Id. at *3-4 (“[H]ad counsel ‘not requested this analysis in order to evaluate a
potential settlement, the [document] and the analysis within it would not have been created.’”).
Similarly, work-product materials do not lose their work-product status because they
incorporate ordinary-course analyses or content.
See id. at *5 (finding financial data
incorporated into work-product document was entitled to the same work product protection
afforded to the entire document because the data was “integral to the privileged purpose of the
document” even when the data, standing alone, would not have otherwise been privileged);
Jud. Watch, Inc. v. DOJ, 432 F.3d 366, 371 (D.C. Cir. 2005) (“[F]actual material is itself
privileged when it appears within documents that are attorney work product.”).12 Accordingly,
the fact that a preexisting analysis by an investment banking firm (over which Google is not
claiming work-product protection) was incorporated as a factual input into Project Sunday
materials, see Opp. Ex. B, Google 30(b)(6) Dep. 116:12-117:6, does not undermine the workproduct nature of the Project Sunday materials. See Mot. 16, 20.
12
See also, e,g., Aerojet Rocketdyne, Inc. v. Glob. Aerospace, Inc., 2019 WL 4929930, at *3
(E.D. Cal. Oct. 7, 2019) (denying in camera review where it was “arguable that the [document]
as a whole was entitled to protection even if every fact or number it contained was not
independently privileged.”); CFPB v. Ocwen Fin. Corp., 2019 WL 1119788, at *3 (S.D. Fla.
Mar. 12, 2019) (preexisting ordinary-course Powerpoint slides were privileged where “the
specific factual information was curated and reformulated by attorneys” and reflected legal
strategy).
12
REDACTED VERSION
PDF Page 14
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 13 of 23 PageID# 2060
Plaintiffs’ argument that the Remedy Projects are not work product because they bear
a “clear nexus to ordinary-course strategic business decisions,” Mot. 15, misstates the
applicable legal standard. The anticipation-of-litigation element hinges on why a document
was created, see Nat’l Union, 967 F.2d at 984, not on whether it has a “nexus” to business
decisions. Under the “because of” test—which applies in this Circuit, id., as well as the D.C.
and Second Circuits—“material generated in anticipation of litigation may also be used for
ordinary business purposes without losing its protected status.” United States v. Deloitte LLP,
610 F.3d 129, 138 (D.C. Cir. 2010); see also United States v. Adlman, 134 F.3d 1194, 1195 (2d
Cir. 1998) (“[A] document . . . does not lose work-product protection merely because it is
intended to assist in the making of a business decision influenced by the likely outcome of the
anticipated litigation.”). As a result, courts in circuits that apply the “because of” test
consistently hold that settlement-evaluation materials are work product, even if they also
inform or lead to business decision-making.13
In the related MDL proceeding pending in the Southern District of New York, Judge
Castel has already deemed references to the substance of one of the Remedy Projects—Project
SingleClick—as protected work product.
He credited a declaration from a Google
13
See FTC v Boehringer (Boehringer I), 778 F.3d 142, 150 (D.C. Cir. 2015) (“We find no
merit in the proposition that any settlement term that has some independent economic value to
both parties must always be treated as an ordinary (non-litigation) business transaction for
purposes of work product protection. Common sense and practical experience teach that
settlement deals routinely include arrangements that could be isolated from the overall
agreement and stand on their own but were nonetheless crafted for the purpose of settling
litigation.”); FTC v Boehringer (Boehringer II), 892 F.3d 1264, 1268 (D.C. Cir. 2018) (holding
that business analysis prepared to assist counsel in evaluating a settlement was work product
even though evaluating the settlement “ultimately was a business decision as well as a legal
decision”); In re Payment Card Interchange Fee & Merch. Disc. Antitrust Litig., 2018 WL
1162552, at *6-11 (E.D.N.Y. Feb. 26, 2018) (holding that settlement workstream documents
“created to assess business implications of the litigation or settlement” and drafted by nonlawyer at the direction of counsel were opinion work product); Bovis Lend Lease, LMB, Inc. v.
Seasons Contracting Corp., 2002 WL 31729693, at *6 (S.D.N.Y. Dec. 5, 2002) (finding
litigation strategy document was still protected by work product doctrine even where it referred
to potential business decisions resulting from the litigation).
13
REDACTED VERSION
PDF Page 15
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 14 of 23 PageID# 2061
businessperson that Project SingleClick was an analysis of actions undertaken in response to
active regulatory investigations, and involved the “subject of liability avoidance” which
reflected the “‘mental impressions, conclusions, opinions, or legal theories of an attorney’
prepared ‘in anticipation of litigation.’” In re Google Digit. Advert. Antitrust Litig., 2023 WL
196146, at *3 (S.D.N.Y. Jan. 17, 2023) (quoting Fed. R. Civ. P. 26(b)(3)). Judge Castel did
not have access to the Google 30(b)(6) deposition testimony regarding the Remedy Projects;
however, the Google declaration that he reviewed explained the project just as the 30(b)(6)
deposition did. Compare id. (citing the declaration as explaining that Project SingleClick was
“an analysis of certain actions undertaken ‘in response to active investigations into, and in
anticipation of litigation concerning, Google’s AdTech business’”), with Opp. Ex. B, Google
30(b)(6) Dep. 144:19-24 (describing Project SingleClick as “an analysis of potential remedies
to some anticipated regulatory actions . . . undertaken in anticipation of litigation concerning
Google’s ad tech business”). And while Judge Castel did not rule that every document referring
to the work of Project SingleClick was categorically protected—as he properly limited his
ruling to the specific documents before him—his reasoning that the work-product doctrine
protects projects undertaken because of anticipated litigation and reflecting legal settlement
strategy applies with equal force here.
By or for a party. The Remedy Project documents also meet the work-product
doctrine’s second element, which asks whether the document was prepared “by or for [the
withholding] party or its representative (including the … party’s attorney, consultant, surety,
indemnitor, insurer, or agent).” Fed. R. Civ. P. 26(b)(3). Plaintiffs cannot dispute that the
Remedy Project documents were created by or for Google.
Instead, Plaintiffs attempt to graft an additional element onto the work-product
doctrine: that the document be prepared “at the direction of an attorney.” Mot. 8 (quoting ePlus
Inc. v. Lawson Software, Inc., 280 F.R.D. 247, 251 (E.D. Va. 2012)). But nowhere does Rule
14
REDACTED VERSION
PDF Page 16
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 15 of 23 PageID# 2062
26(b)(3) require that the document be prepared by or at the direction of an attorney. See Hempel
v. Cydan Dev., Inc., 2020 WL 4933634, at *4 (D. Md. Aug. 24, 2020) (“[T]he statutory text of
Rule 26 does not require that documents be ‘prepared by, for, or at the request of [a] lawyer’
in order to be protected as work product.”). The work-product doctrine, as the Fourth Circuit
has recognized, protects documents prepared in anticipation of litigation “whether the material
was actually prepared by [an] attorney or by another ‘representative’ of the [withholding]
party.” Nat’l Union, 967 F.2d at 984.14
As explained in the accompanying Lazarus Declaration, each of the Remedy Project
documents was either prepared in response to active regulatory investigations or reveals
substantive details of the Remedy Projects. See Opp. Ex. A, Lazarus Decl. ¶¶ 5-9. And as
explained in that Declaration, Google’s in-house counsel initiated, oversaw, and/or played a
key role in each of the Remedy Projects. See id.; see also Opp. Ex. B, Google 30(b)(6) Dep.
74:19-75:5, 88:14-15, 122:16-123:18, 128:11-129:9, 146:11-13, 149:7-12, 151:5-17, 182:16-
14
As this Court and others have recognized, work product extends to materials beyond those
prepared by, or at the direction of, attorneys. See Dunston v. Cecil Huang, 2010 WL 11698098,
at *1 (E.D. Va. May 5, 2010) (Anderson, M.J.) (holding that material “created by a
representative of the defendants” in anticipation of litigation was “protected by the work
product protection”); see also, e.g., Williams v. AT&T Mobility, 2022 WL 2821922, at *5
(E.D.N.C. July 19, 2022) (rejecting argument that “since the documents were created by nonlawyers, they cannot be privileged [as work product]”); Coffey v. Tyler Staffing Servs., Inc.,
2020 WL 1307233, at *3 (W.D. Va. Mar. 19, 2020) (“The fact that these materials were not
prepared by an attorney does not, by itself, disqualify them from falling under the work product
doctrine.”); Gutshall v. New Prime, Inc., 196 F.R.D. 43, 46 (W.D. Va. 2000) (finding materials
that party itself “prepared, or commissioned the preparation of” constituted work product);
Haigh v. Matsushita Elec. Corp. of Am., 676 F. Supp. 1332, 1356-58 (E.D. Va. 1987) (rejecting
defendants’ argument that tape recordings were not work product because they were made
absent attorney direction, explaining that “Defendants are living in the past, and are presenting
a pre-1970 argument”—i.e., from before Rule 26(b)(3) was amended). To the extent that ePlus
suggests otherwise, Mot. 8, it contradicts the plain language of Rule 26(b)(3). Relying solely
on Hickman v. Taylor, 329 U.S. 495 (1947), see ePlus, 280 F.R.D. at 257, the ePlus court failed
to consider that the amendments to Rule 26(b)(3) expanded the work-product doctrine. Moore
v. Tri-City Hosp. Auth., 118 F.R.D. 646, 649 (N.D. Ga. 1988) (“Rule 26(b)(3) . . . preserves
the essential portions of the doctrine as announced in Hickman, but also notably expands the
doctrine by extending discovery protection to the work product of a party as well as that party’s
attorneys.”).
15
REDACTED VERSION
PDF Page 17
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 16 of 23 PageID# 2063
183:1, which further corroborates that the Remedy Projects were undertaken in anticipation of
litigation and reflect legal strategy. This is analogous to Zetia, where the court found potential
settlement analyses to be work product in part because they were initiated by and overseen by
in-house counsel. See 2019 WL 6122012, at *3.
Plaintiffs mischaracterize Google’s work-product assertions over Remedy Project
documents as shielding “major business projects simply because they were undertaken while
the company was under government investigation.” Mot. 7. But Google has consistently made
clear, through extensive correspondence with DOJ,15 the February 28, 2022 deposition of its
corporate representative, and a meet-and-confer with DOJ in May of that year, that the basis
for its assertion of work-product protection for the Remedy Project documents is that they were
created because of active government investigations and anticipated litigation.16 This is exactly
what the law requires for the work-product protection to apply. Each of the 17 Remedy Project
documents was either created as part of or reveals the substance of at least one of the Remedy
15
See, e.g., Mot. Ex. 29, Ltr. from J. Elmer, Feb. 25, 2022.
16
Although one of the Remedy Projects—Project Banksy—involved the analysis, for the
purpose of potentially settling a regulatory investigation, of a concept sourced in a nonprivileged, prior workstream referred to simply as “Banksy,” that does not preclude Project
Banksy from receiving work-product protection. Project Banksy involved the analysis and
eventual execution of some of the ideas from non-privileged “Banksy” in the specific context
of the French Competition Authority and other regulators’ investigations. See Opp. Ex. C,
Dep. of
at 168:1-172:10 (describing “Banksy”); Opp. Ex. B, Google 30(b)(6)
Dep. 200:22-201:7 (explaining Project Banksy as emerging from “Banksy” in response to the
French Competition Authority’s investigation). That analysis and execution may not have been
undertaken at all, and certainly would not have been undertaken in substantially similar form,
but for the French Competition Authority investigation and Google’s legal strategy to identify
settlement options in response to that investigation. See Opp. Ex. B, Google 30(b)(6) Dep.
210:15-211:6. Moreover, Google has asserted privilege over only Project Banksy materials,
but not prior “Banksy” materials, which underscores that Google has properly claimed workproduct protection over only those documents reflecting the company’s legal strategy and
where the active regulatory investigations and anticipated litigation were the driving force
behind the documents’ creation.
16
REDACTED VERSION
PDF Page 18
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 17 of 23 PageID# 2064
Projects, in whole or in part, see Opp. Ex. A, Lazarus Decl. ¶¶ 11-15, and has therefore been
properly redacted or withheld.
III.
Google Has Properly Redacted or Withheld Attorney-Client Privileged
Documents
The remaining two documents at issue concern the attorney-client privilege.17 The
singular basis for Plaintiffs’ challenge to these documents is Google’s “Communicate with
Care” training. According to Plaintiffs, this training caused some Google employees to label
communications privileged and confidential and include an attorney even when the
communication was not privileged. Mot. 11-12. From there, Plaintiffs speculate that some of
the documents Google has withheld in this case must have been the product of “Communicate
with Care,” rather than of good-faith privilege determinations. There is no basis for such
speculation. As Google has explained to Plaintiffs, Google has not claimed privilege based on
“Communicate with Care,” nor has Google withheld documents on the basis of privilege
simply because they contain the words “Privileged and Confidential.” Google has reviewed
documents one by one and has made good faith privilege determinations based on the substance
of each document even within a voluminous and fast paced production. A document Plaintiffs
appended as an exhibit to their Motion, see Mot. Ex. 5, underscores Google’s focus on a
document’s substance—and not whether it is labeled privileged. See Mot. 12 (observing
produced document appended as Exhibit 5 contained a privilege legend, but lacked “any
request for legal advice”).
Plaintiffs’ challenges to these two documents do not reflect any pervasive issues with
Google’s privilege log or Google’s application of the attorney-client privilege:
17
These two documents—excluding the ones that Google has thrown back, see supra n.2—are
Documents 3 (GOOG-DOJ-AT-00029680) and 7 (GOOG-DOJ-AT-01139556).
17
REDACTED VERSION
PDF Page 19
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 18 of 23 PageID# 2065
Document 3 (GOOG-DOJ-AT-00029680). Google has in good faith lifted the second
redaction at -685 and the redaction at -686. The document now contains a single redaction at
-685, containing feedback on the deal terms from “Legal,” which “contain[s] legal advice of
ESQ regarding contract terms,” as noted in the privilege log.
Document 7 (GOOG-DOJ-AT-01139556). This is a 47-page presentation titled
“Streamline programmatic buying doors on AdMob and Ad Manager,” containing redactions
at -564, -569, 570, -579, -589, and -602.18 The redacted portions on -564, -579, and -602 are
comments seeking, containing, and responding to legal advice given by Google’s in-house
counsel
about information on the previous slide (-563, and duplicated at -578
and -601), detailing revenue share scenarios for AdMob and Ad Manager. The redaction at
-569 contains “[f]eedback from legal”—plainly facially privileged material—and the redacted
rows on -570 convey “[r]egulatory [r]isk” and potential mitigating actions reflecting legal
advice of
.
In addition, while Plaintiffs take issue with the privilege-log
description for these redactions, they do not explain why that description—“seeking and
revealing legal advice of
ESQ regarding legal aspects of product
development”—is “materially deficient.” Mot. 24.19
Regarding the redaction at -589, after re-reviewing this document in good faith, Google
has removed this redaction of a single slide within a 47-page presentation and has provided
Plaintiffs with a replacement copy of this document, see supra n.2.
18
The three pages -564, -579, and -602 are exact copies of each other and the redactions were
identical except for a portion at the bottom of -564, which had been inadvertently left
unredacted. Google has taken steps to conform the redactions on -564 to the redactions on -579
and -602 and has reproduced the document to Plaintiffs accordingly.
19
Plaintiffs claim that no lawyer appears in the metadata for this document. That claim is
incorrect, as there are eight attorneys in the presentation’s metadata, including
and outside counsel.
18
REDACTED VERSION
PDF Page 20
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 19 of 23 PageID# 2066
Documents 4 (GOOG-DOJ-AT-01019463) and 21 (GOOG-DOJ-06588656). Google
has reviewed these two documents and has in good faith opted to drop its claims of privilege
over these two documents. As with the voluntarily lifted targeted redactions in Documents 3,
7, 15, and 18, lifting the redactions from Documents 4 and 21 reflects Google’s consistent and
ongoing good faith and reasonable efforts to balance the production of relevant material with
zealously protecting material that is legally privileged. See Transcript of Status Conf. at 62,
United States v. Google LLC, No. 1:20-cv-03010-APM (D.D.C. Oct. 20, 2020), ECF No. 353
(viewing Google’s throwbacks as part of a “good-faith effort to . . . re-review [withheld
documents], re-categorize them, de-privilege those that need to be de-privileged, and produce
them to the plaintiffs”); Google Digit. Advert., 2023 WL 196146, at *3 (Castel, J.) (“The Court
declines to ascribe any nefarious motive to [Google’s] change of position” by no longer
asserting privilege over a challenged document); see also Am. Nat’l Bank & Tr. Co. of Chi. v.
Equitable Life Assurance Soc’y of U.S., 406 F.3d 867, 870, 879 (7th Cir. 2005) (defendant
“exhibited good faith throughout the privilege log proceedings [by] among other acts,” rereviewing and throwing back documents as necessary).
IV.
Plaintiffs Improperly Seek an Advisory Opinion
The Court should decline to grant Plaintiffs the advisory opinion they seek. Although
Plaintiffs contend that they have selected a “representative” sample of documents, Mot. 1, they
have cherry-picked 21 documents out of nearly 180,000 documents withheld or redacted over
the course of a three-year investigation, each dissimilar and covering a range of issues, from
the millions of documents produced by Google. From this small and cherry-picked sample, it
is not possible to extrapolate any conclusions about Google’s overall production and privilege
log.
Plaintiffs’ request that the Court treat a review of 21 documents as “exemplars of three
categories of documents,” Mot. 4, is not how privilege determinations work. Rather, privilege
19
REDACTED VERSION
PDF Page 21
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 20 of 23 PageID# 2067
review involves “a document-by-document privilege analysis.” Interbake, 637 F.3d at 503
(“[E]ach e-mail within a particular line of discussion must be analyzed separately for privilege
purposes.”); see also In re Rivastigmine Pat. Litig., 2005 WL 2319005, at *5 (S.D.N.Y. Sept.
22, 2005) (declining to rule on privilege not tied to “specific documents” because “privilege
determination is fact-sensitive”). In the parallel MDL proceeding, Judge Castel rejected a
similar request by those plaintiffs to “clarify” how the attorney-client privilege would apply to
documents not submitted for in camera review, concluding that it was “essentially a request
for an advisory opinion on legal principles.” Google Digit. Advert., 2023 WL 196146, at *1.
The same reasoning applies equally here.
V.
Google’s Exercise of Its Clawback Rights Does Not Justify In Camera Review
Plaintiffs try to use the fact that Google clawed back documents to argue that the Court
should review the 21 documents in camera. But the fact that Google has exercised its right to
claw back inadvertently produced privileged material has no bearing on whether in camera
review is appropriate. See supra Point I. Plaintiffs cite no case law suggesting otherwise.20
VI.
Google Has Not Waived Privilege over Document 1 (GOOG-DOJ-12766025).
Plaintiffs contend that Google has waived privilege over one document (GOOG-DOJ-
12766025),21 by failing to object or claw back material read into the record during a Google
20
Plaintiffs cannot use the fact that Google has clawed back documents as a backdoor to in
camera review. Plaintiffs say that because “all of the documents identified in the Appendix
have been clawed back at one time or another,” in camera review will help them “articulate
fully their challenge to Google’s claims of privilege.” Mot. 24. Not only is that not the standard
for in camera review, see supra Point I, but it also makes little sense. It is, of course, true
across all privilege challenges—whether privilege was asserted along with an initial production
or asserted by clawback—that the opposing party has a limited ability to understand the
withheld content and make arguments against privilege. But if the challenging party could
force in camera review simply by claiming that it needed to see materials withheld as privileged
in order to challenge even more withheld materials, then the in camera standard would be
meaningless.
21
In Document 1 (GOOG-DOJ-12766025), Google redacted communications revealing the
substance of Projects SingleClick and Stonehenge. Because Projects SingleClick and
Stonehenge refer to Remedy Projects initiated because of anticipated litigation, and producing
20
REDACTED VERSION
PDF Page 22
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 21 of 23 PageID# 2068
employee’s August 10, 2021 deposition. See Mot. 19 n.33. But Plaintiffs mischaracterize the
record. Google did not allow portions of the document to be read into the record at the
deposition without objection. Instead, DOJ read a portion of the document into the record as
part of a question, and before counsel could instruct the witness not to answer and claw back
the document, the witness responded. See Mot. Ex. 18, Dep. of
at
197:17-23. When DOJ sought to clarify the witness’s answer, counsel for Google objected.
Id. at 197:24-198:8. After a short break, DOJ restated its question, and counsel for Google
instructed the witness not to answer on the basis of privilege. Id. at 198:16-199:3. The
document read into the record was then immediately clawed back as privileged during the
deposition. Id. at 199:4-11, 201:22-202:2.
These facts make it clear that Google did not waive privilege. On the contrary,
Google’s counsel immediately informed DOJ during the deposition that the exhibit contained
privileged material with respect to Projects Stonehenge and SingleClick, and promptly took
steps to rectify the disclosure by clawing back the document and instructing the witness not to
answer additional questions. These were reasonable steps to protect privilege in light of an
inadvertent production. See, e.g., King Pharm., Inc. v. Purdue Pharma, L.P., 2010 WL
2243872, at *2 (W.D. Va. June 2, 2010) (“Purdue clearly acted promptly in taking reasonable
steps to rectify the error” when it clawed back, during a deposition, “one of millions of
documents produced by the parties in [a] complex [] case.”).22
Moreover, the Google
employee, in his individual capacity, could not waive privilege or work-product protection on
the document unredacted would reveal Google settlement strategy and analysis, the material is
properly redacted as work product. See supra Point II.
22
See also Karma Auto. LLC v. Lordstown Motors Corp., 2021 WL 4147007, at *3 (C.D. Cal.
June 16, 2021) (holding that privilege was not waived where counsel “object[ed] on the record
at [witness’s] deposition and request[ed] that [non-disclosing party] destroy and refrain from
further use of the document”).
21
REDACTED VERSION
PDF Page 23
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 22 of 23 PageID# 2069
behalf of the company by answering a question before the company’s counsel had time to
object. “[T]he issue of waiver in the case of the testimony of a corporation’s agent is whether
the corporation has acted deliberately in disclosing privileged information, not whether the
witness acted deliberately.” Parneros v. Barnes & Noble, Inc., 332 F.R.D. 482, 504-05
(S.D.N.Y. 2019) (emphases added). And nothing in the record suggests that Google intended
to waive privilege. Google has consistently taken the position that documents relating to the
Remedy Projects are protected work product.
CONCLUSION
For the foregoing reasons, the Court should deny Plaintiffs’ Motion.
Dated: June 2, 2023
Respectfully submitted,
Eric Mahr (pro hac vice)
Andrew Ewalt (pro hac vice)
Julie Elmer (pro hac vice)
Lauren Kaplin (pro hac vice)
Jeanette Bayoumi (pro hac vice)
Claire Leonard (pro hac vice)
Sara Salem (pro hac vice)
Tyler Garrett (VSB # 94759)
FRESHFIELDS BRUCKHAUS
DERINGER US LLP
700 13th Street, NW, 10th Floor
Washington, DC 20005
Telephone: (202) 777-4500
Facsimile: (202) 777-4555
eric.mahr@freshfields.com
/s/ Craig C. Reilly
Craig C. Reilly (VSB # 20942)
THE LAW OFFICE OF
CRAIG C. REILLY, ESQ.
209 Madison Street
Alexandria, VA 22314
Telephone: (703) 549-5354
Facsimile: (703) 549-5355
craig.reilly@ccreillylaw.com
Daniel Bitton (pro hac vice)
AXINN, VELTROP & HARKRIDER LLP
55 2nd Street
San Francisco, CA 94105
Telephone: (415) 490-2000
Facsimile: (415) 490-2001
dbitton@axinn.com
Bradley Justus (VSB # 80533)
Koren Wong-Ervin (pro hac vice)
AXINN, VELTROP & HARKRIDER LLP
1901 L Street, NW
Karen L. Dunn (pro hac vice)
Jeannie H. Rhee (pro hac vice)
William A. Isaacson (pro hac vice)
Joseph Bial (pro hac vice)
Amy J. Mauser (pro hac vice)
Martha L. Goodman (pro hac vice)
Bryon P. Becker (VSB #93384)
Erica Spevack (pro hac vice)
PAUL, WEISS, RIFKIND, WHARTON &
GARRISON LLP
2001 K Street, NW
Washington, DC 20006-1047
Telephone: (202) 223-7300
Facsimile (202) 223-7420
kdunn@paulweiss.com
Meredith Dearborn (pro hac vice)
22
REDACTED VERSION
PDF Page 24
Case 1:23-cv-00108-LMB-JFA Document 245 Filed 06/02/23 Page 23 of 23 PageID# 2070
Washington, DC 20036
Telephone: (202) 912-4700
Facsimile: (202) 912-4701
bjustus@axinn.com
PAUL, WEISS, RIFKIND, WHARTON &
GARRISON LLP
535 Mission Street, 24th Floor
San Francisco, CA 94105
Telephone: (646) 432-5100
Facsimile: (202) 330-5908
mdearnborn@paulweiss.com
Counsel for Defendant Google LLC
23
REDACTED VERSION
Back to 
