Three separate judges recently remarked on Google’s failure to preserve and produce documents relevant to antitrust claims against Google. (More here.) In response, critics asked the State Bar of California to investigate Google General Counsel Kent Walker for violating the California Professional Code of Conduct in counseling Google to destroy records to multiple ongoing lawsuits, including his 2008 memo instructing employees to “avoid inadvertent retention” of communications that might be “used against” Google in litigation.
From the timeline, one might think Google’s document preservation practices were perfect until the recent matters. Note memo written in 2008 versus litigation in 2023-2024 questioning document retention. If there were no discovery grievances during that fifteen-year period, maybe this is the first time Google failed to conduct itself properly during discovery – or at least the first time anyone figured out the problem and called Google out on it. If so, maybe this is genuinely some kind of mistake, and maybe the sanction on Google should be correspondingly reduced.
Not so. I have been following Google litigation for longer than most. When I read the recent briefs and decisions about Google’s failure to preserve documents, I immediately thought of a different episode, a decade and a half ago, when Google failed to preserve documents relating to YouTube despite knowing about likely litigation.
Google’s failure to preserve documents relating to copyright infringement at YouTube
When Google considered acquiring YouTube, copyright concerns loomed large. Analyzing a sample of videos from YouTube, Credit Suisse bankers found that more than 60% of YouTube video views were “premium” (their euphemism for copyright), and only 10% of those were licensed. In this period, YouTube faced competition from, among others, Google Video—which took genuine steps to protect copyright. Seeing YouTube traffic soar while they struggled, Google Video staff were unsparing in their criticism of YouTube’s approach to copyright, calling YouTube, “a rogue enabler of content theft” “completely sustained by pirated content” “trafficking mostly illegal content”. (See also Google’s slides about copyright risks and rights-holder criticism.) Concerned about a “precedent-setting lawsuit” about copyright, Google held back a 12.5% escrow from the YouTube acquisition price to cover possible copyright liability. With Google’s own staff explicitly criticizing YouTube’s approach to copyright, and Google’s term sheet discussing copyright risks, copyright litigation was not just foreseeable; it was actually foreseen, then measured and priced.
Under longstanding law, when litigation is reasonably foreseeable, a party is obliged to preserve documents for use as evidence. (See e.g. Silvestri v. General Motors, 271 F.3d 583, 590.) From the start, copyright infringement at YouTube was far beyond that modest standard. So from the first moment Google began to consider acquiring YouTube, Google should have preserved all relevant documents. Certainly when the acquisition was completed, with copyright concerns still looming large and threats from rights-holders ongoing, Google was obliged to preserve.
Rather than preserve documents relating to copyright infringement at YouTube, Google did much the opposite. The most brazen missing emails are those sent and received by then-CEO Eric Schmidt. In a May 6, 2009 deposition, Schmidt indicated having about 30 computers (page 7 line 10) but configuring them “to not retain my e-mails unless asked specifically” (17:9-10). Asked whether he deleted them manually versus through some sort of automation, Schmidt said “the answer would vary” (18:17-18). He said he would “delete or otherwise cause the emails that I had read to go way as quickly as possible” (18:24 to 19:2). Asked whether he began to preserve documents after a copyright lawsuit was filed alleging widespread infringement at YouTube, Schmidt said “I did change my practice after this lawsuit was filed and I was notified” but refused to reveal details which he indicated were attorney-client privileged. Despite the possible change to Schmidt’s retention practice after the Viacom case was filed, searching his email for responsive documents yielded only 19. (Hohengarten declaration ¶266) Google’s acquisition of YouTube was its largest transaction to date. It is inconceivable that the CEO of Google would have just 19 emails pertaining to its largest acquisition. By his own admission, Schmidt “deleted or otherwise cause[d]” to be deleted the huge number of otherwise-responsive emails. Thousands? Tens of thousands? Meanwhile, if he changed his document retention practice after the lawsuit was filed, why aren’t there more documents from that period? He wouldn’t say, and it seems the world will never know.
In principle, excellent memory could stand in for the deleted emails. The Google executive triumvirate—Schmidt, Page, and Brin—had no shortage of intellectual firepower, and arguably they’d be expected to have correspondingly excellent memories. In fact, Business Insider specifically credits co-founder Larry Page with a “photographic memory.” Yet in his deposition about the YouTube acquisition, Page said he could not remember YouTube’s copyright procedures, whether Google wanted an indemnification from YouTube for copyright liability, whether Google executives discussed infringing content at YouTube, whether he reviewed due diligence about the acquisition of YouTube, whether he attended the board meeting when the Google board approved acquiring YouTube, or even whether he was in favor of the purchase or against it. To these questions among many others, Page remarked “I don’t recall” and the like, over and over — by my count, 132 times. See the questions and verbatim quotes of Page’s answers.
(Incidentally, even Page’s deposition transcript almost didn’t see the light of day. Initially, this deposition was filed under seal in Viacom v. Google litigation as Hohengarten declaration part 11 (docket #222) attachment #20, redacted in full at Google’s request. See also docket #247, ordering that redaction. But in rechecking the docket incidental to this article, I found later docket #324, which provides the transcript. Last month I uploaded these files to Courtlistener for public archival and review. Page’s deposition is Attachments 5, 6, and 7. Until I uploaded Page’s deposition transcript to Courtlistener, these files lingered in PACER – unindexable by web search, available only for those who knew exactly where to look, knew how to log in, and were willing to pay per-page access fees.)
The memory of YouTube co-founder Chad Hurley was equally poor, but newly revealed in Attachments 2-4 in the same docket #324. Hurley had no difficulty recalling most aspects of product strategy, negotiation strategy, and distribution strategy. However he was unable to recall multiple subjects that were sensitive in light of the litigation: the basis of Google’s valuation of YouTube (part 1 pages 29-32), entire movies uploaded to YouTube (part 2 pages 12-16), copyright enforcement policies (part 2 pages 19-20 and 26-30), and why YouTube removed ads from Watch Pages (part 2 pages 23-24)—Hurley said he couldn’t recall.
Separately, Viacom calls out Google’s failure to produce relevant documents within its possession. For example, on March 22, 2006, YouTube founder Jawed Karim presented to the YouTube board of directors:
As of today episodes and clips of the following well-known shows can still be found: Family Guy, South Park, MTV Cribs, Daily Show, Reno 911, Dave Chapelle. This content is an easy target for critics who claim that copyrighted content is entirely responsible for YouTube’s popularity. Although YouTube is not legally required to monitor content (as we have explained in the press) and complies with DMCA takedown requests, we would benefit from preemptively removing content that is blatantly illegal and likely to attract criticism.
Having acquired YouTube, Google should have received the materials previously presented to YouTube’s board. Yet Google failed to produce this document to Viacom in litigation. Viacom only received it when Karim, to his credit, produced it after he left YouTube.
Viacom’s Motion for Summary Judgment criticizes Schmidt’s failure to preserve relevant documents, as well as what it calls witnesses’ “serial amnesia” and Google’s other failures to produce. But Viacom never filed a motion seeking sanctions for spoliation. The court ultimately granted Google’s motion for summary judgment without ruling on Google’s alleged failure to preserve, witnesses too conveniently forgetful, or any other discovery violation.
Rereading Google’s deficient discovery in the Viacom’s matter, in conjunction with 2023-2024 proceedings about Google discovery violations, to me the picture became increasingly clear. Google’s decision to evade discovery obligations isn’t a new problem, and it isn’t really a change. Rather, this problem has been lingering for years, and has surely harmed scores of other litigants—whose ability to prove their cases against Google was stymied by Google falling short of both the productions and the candor required by court rules. For broader thoughts and more on the other discovery disputes, see Revisiting litigation alleging Google discovery violations.