Before: Chin and Carney, C.JJ., and Sannes, D.J.1 ON APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF CONNECTICUT This is an appeal from a post-trial ruling by the United States District Court for the District of Connecticut (Shea, J.) granting in part the request of Lawrence Dressler, a nonparty, for release of the video recording of the deposition of Aviad Hack, also a non-party witness. In that deposition, Hack testified that, while a minor, he was a victim of sexual abuse by the defendant, Daniel Greer. Hack also admitted to having become aware of Greer’s sexual abuse of the plaintiff, Eliyahu Mirlis, when Hack was an adult and Mirlis was a minor. The District Court ruled that the portions of the deposition video that had been shown to the jury during trial in place of Hack’s live testimony were judicial documents subject to a strong presumption of public access. The District Court further determined that Hack’s privacy interest in the deposition video was insufficient to rebut that presumption. The court therefore ordered that the portions of the video-recorded deposition that were shown to the jury be released to Dressler and the public, notwithstanding the public availability of a transcript of those same portions of the deposition. Hack now challenges the District Court’s ruling. On review, we conclude that the District Court erred by failing to take into account Dressler’s motives in obtaining, and likely course of action with, the video recording. We further conclude that the District Court accorded insufficient weight to Hack’s privacy interests as a minor victim. REVERSED. SUSAN CARNEY, C.J. Appellant Aviad Hack (“Hack”) was a non-party witness in a 2016 civil lawsuit in the District of Connecticut. In that suit, the plaintiff, Eliyahu Mirlis (“Mirlis”), a former student at a religious school in New Haven, Connecticut, accused the defendant, Daniel Greer (“Greer”), the former religious leader of that school, of sexually abusing Mirlis when Mirlis was a minor and a student there. Mirlis sought damages from Greer for that abuse. In connection with those proceedings, Hack gave a video-recorded deposition lasting several hours in which he testified in detail to having also been a victim of Greer’s sexual abuse several decades earlier, when Hack was a minor and a student at the school. Hack further testified in his deposition that he had known about Greer’s abuse of Mirlis, which occurred when Hack was an adult and employed at the school. A transcript of the deposition is publicly available, but public access to the video recording itself is the subject of this appeal. Hack voluntarily sat for the deposition in 2016, but he did not voluntarily appear at trial in 2017; indeed, at the school in Rhode Island where he was teaching in 2017, he successfully evaded those who were attempting to serve a subpoena on him to appear at the civil proceedings in Connecticut. Accordingly, in place of his live testimony, the United States District Court for the District of Connecticut (Shea, J.) permitted the parties to play portions of Hack’s deposition video for the jury. After trial, Lawrence Dressler (“Dressler”), the Interested Party-Appellee in the appeal now before us, sought the District Court’s leave to access the video recording of Hack’s entire deposition. Using the screenname “Larry Noodles,” Dressler had written voluminously on his blog about the trial and disparagingly about both Hack and Greer. He informed the court that he sought to copy the video so that he could post it publicly to his internet blog. Hack vigorously opposed Dressler’s motion. His counsel argued that “the privacy interests of Mr. Hack in not having the video of his deposition distributed in public forums and replayed to the masses[] far outweighs the need [for] access [to the] same.” Mot. for Protective Order at 1, Mirlis v. Greer, No. 16-cv-678 (D. Conn. Jan. 2, 2018), ECF No. 256. While this motion was pending, on March 8, 2018, an unsealed transcript of Hack’s deposition, edited to show only the portions of Hack’s deposition that were played for the jury, was filed with the court and noted on the District Court docket.2 It was made, and is still, available for public examination. The District Court determined that those portions of the video-recorded deposition that had been shown to the jury at trial constituted a “judicial record” subject to a strong common law presumption of public access under Circuit precedent. Then, relying primarily on our 1987 decision in Application of CBS, Inc., 828 F.2d 958, 959 (2d Cir. 1987) (“CBS”), the District Court further determined that Hack’s privacy interests in those excerpts of the deposition video were not sufficient to override the presumption of public access. The court reasoned that, although the video captured Hack describing Greer’s predatory sexual assault on Hack when Hack was a minor, the video’s subject matter was less “gruesome” than the content of videos in cases where public access has been denied, and therefore, under standards derived largely from CBS, it was less compellingly private. Hack App’x 143. The District Court also found persuasive the argument that Hack’s privacy interests in the portions of the video that had been played for the jury were reduced because a transcript of the entire deposition was already publicly available. It therefore ordered that Dressler be given access to a video of those portions of the deposition, to do with as he liked. Recognizing the sensitivity of its decision, however, the court stayed its order pending appeal. On review, we conclude that the District Court committed reversible error in determining that those portions of the video-recorded deposition that had been played for the jury should be publicly released for copying and general display on the Internet. We agree with the District Court’s initial conclusion that, having been displayed to the jury during trial, those portions of the video were “judicial documents” as to which a presumptive right of public access applied. But, in balancing the weight of the presumption of access, the District Court failed to take sufficient account of two categories of countervailing considerations: (1) Dressler’s motive for obtaining, and intent in releasing, the deposition video; and (2) Hack’s privacy interests. First, the court erred as a matter of law by failing to give any consideration to Dressler’s apparently spiteful motives and announced course of action with the footage. Although we have commented generally that motive is “irrelevant to defining the weight accorded the presumption of access,” United States v. Amodeo, 71 F.3d 1044, 1050 (2d Cir. 1995) (“Amodeo II”),3 the Supreme Court has instructed that access to judicial documents should not be permitted “to gratify private spite or promote public scandal with no corresponding assurance of public benefit,” Nixon v. Warner Commc’ns, Inc., 435 U.S. 589, 603 (1978) (“Nixon”), a limitation that we find relevant here. Second, the District Court undervalued the weight properly accorded the intense intrusion on Hack’s privacy interests that the internet publication of the video excerpts would effect. Today, unlike in the era of our decision in CBS, videos of all types are routinely and widely shared on the Internet, where (as far as we can predict now) it appears they will be available in perpetuity for unlimited viewing, further dissemination, and easy manipulation; their subjects are unable to escape them. Yet Dressler identified no affirmative public need that would be served by general release of the video when the transcript was already available. Indeed, no other individual or entity, and no media outlet, joined in Dressler’s request. The privacy interests of a third-party witness in a deposition video in which that witness describes being the minor victim of a sex crime perpetrated by a trusted adult are entitled to great weight and sensitive consideration in the court’s balancing of the type of access sought against the privacy rights asserted. Accordingly, we REVERSE the order of the District Court and REMAND the cause with instructions that the District Court enter an order denying Dressler’s request for the video recording. BACKGROUND In May 2016, Mirlis sued the Yeshiva of New Haven, Inc. (the “Yeshiva”) and Greer, the former principal and rabbi of the Yeshiva, alleging that Greer sexually abused Mirlis in the early 2000s, when Mirlis was in his teens and a student at the Yeshiva, and Greer was in his 50s. Mirlis sought damages for the harm to him. In 2017, the case went to trial and the jury returned a verdict in favor of Mirlis, awarding him $21.7 million in compensatory and punitive damages. Greer’s appellate challenges to the trial proceedings and the damages award are the subject of a separate opinion published concurrently with this writing.4 In preparation for trial, on July 25, 2016, and August 2, 2016, Mirlis conducted a video-recorded deposition of Hack as a non-party fact witness with knowledge relevant to the civil litigation. Hack, by that time in his 40s and a resident of Rhode Island, served as assistant dean of the Yeshiva during Mirlis’s time as a student there. During his deposition, Hack testified that he, too, had been a victim of Greer’s sexual abuse when he was a student at the Yeshiva in the late 1990s, several years before Mirlis arrived. Hack also admitted that he became aware when he was assistant dean of the Yeshiva that Greer was assaulting Mirlis, and made no report to state authorities. Although Hack willingly participated in the deposition, he went to great lengths later to evade process servers who were attempting to subpoena him to testify at trial. Three process servers submitted affidavits to the District Court, one averring that, when he approached Hack at the school in Rhode Island where Hack then taught, Hack “ran out of his classroom, down the hallway[,]…and out the back door of the school, leaving behind a classroom full of students.” App’x 442. The other affidavits offered similar sworn statements. For two weeks after the service attempts began, Hack called in sick to the school. Despite surveilling the school and his home, no process server was able to serve the subpoena, and Hack did not appear at trial. Based on these affidavits, the District Court ruled at Mirlis’s request that Hack was an “unavailable witness” within the meaning of Rule 32(a)(4) of the Federal Rules of Civil Procedure and that, accordingly, Mirlis would be permitted to introduce at trial relevant portions of Hack’s deposition video. Mirlis did so. In those deposition excerpts, Hack described a sexual relationship that Greer initiated with him in the early 1990s, when Hack was about sixteen years of age and a student at the Yeshiva, and Greer was the school’s rabbi and leader. Hack further averred that the relationship continued through the years after 2000, when Hack became a teacher at the Yeshiva, and stopped sometime in 2004, when Hack married. The jury also saw portions of the deposition in which Hack admitted that, in 2003, he learned from Greer himself that Greer was engaging in sexual contact with Mirlis, who was then a minor and a student at the Yeshiva, as Hack had been when his relationship with Greer began. After four days of trial in May 2017, the jury returned a verdict in favor of Mirlis. Several months after the verdict issued, non-party and non-witness Dressler, a New Haven resident who had attended the trial and posted about it on his “Larry Noodles” blog, wrote to the District Court, requesting that the court make public and available for reproduction the entirety of Hack’s video-recorded deposition. On the blog, Dressler wrote generally about criminal justice issues and prison life (he had been incarcerated from May 2014 to November 2015). He also had written extensively about Greer’s civil trial on Mirlis’s claims — coverage that he claimed had attracted hundreds of thousands of views. Hack, citing his privacy concerns and the sensitive nature of the material, opposed Dressler’s request and sought a protective order that would prohibit release of the video recording of his deposition. (Greer also opposed Dressler’s request.) After briefing and oral argument, in January 2018 the District Court granted Dressler’s request for the release for copying and redistribution of those portions of the video that were shown to the jury at trial. At the District Court’s direction, a transcript of the deposition, edited to omit the portions of Hack’s testimony not played for the jury, was posted on the court’s docket in March 2018. Hack timely appealed the District Court’s denial of his motion for a protective order. The District Court stayed execution of its disclosure order pending this Court’s review. DISCUSSION Our Court has not articulated with precision the standard of review that governs appeals from a district court’s decision to grant an intervenor’s request to copy audio and visual recordings adduced in court proceedings. See United States v. Graham, 257 F.3d 143, 148 (2d Cir. 2001). On review of a district court’s order to seal or unseal documents, an order in many respects analogous to the present situation, “we examine the court’s factual findings for clear error, its legal determinations de novo, and its ultimate decision to seal or unseal for abuse of discretion.” Brown v. Maxwell, 929 F.3d 41, 47 (2d Cir. 2019). In Nixon, the Supreme Court recognized that the presumptive public right of access to judicial documents may include access for purposes of copying audio recordings. 435 U.S. at 615-16. We observed in Graham, however, that the Nixon decision was “ambiguous” as to “whether we should review only for abuse of discretion, or whether a more searching review is warranted.” 257 F.3d at 148-49. We ultimately left “for another day” the choice between abuse-of-discretion review and a “more searching review” because in Graham, we would have affirmed the district court’s decision to grant such access under either standard. Id. Hack and Dressler agree that the abuse-of-discretion standard should apply, but we think the question both unsettled and potentially significant in similar cases because of the weighty interests that may be involved. Again, however, because the standard of review that we apply here does not determine the result — reversal of the District Court’s decision to release portions of the deposition video to Dressler — we once again defer any refinement of the “more searching” standard and any decision as to which standard should govern. I. Common Law Presumption of Public Access to Judicial Documents Judicial documents are subject at common law to a potent and fundamental presumptive right of public access that predates even the U.S. Constitution.5 United States v. Amodeo, 44 F.3d 141, 145 (2d Cir. 1995) (“Amodeo I”); see also Lugosch v. Pyramid Co., 435 F.3d 110, 119 (2d Cir. 2006) (describing right of public access as “firmly rooted in our nation’s history”). That right includes “a general right to inspect and copy” such judicial documents. Nixon, 435 U.S. at 597. To determine whether the presumption attaches to a particular record, our Court engages in a three-step inquiry. See Lugosch, 435 F.3d at 119-20. First, the court determines whether the record at issue is a “judicial document” — a document to which the presumption of public access attaches. Id. at 119. Not all documents filed with the court are “judicial” documents. Rather, a judicial document is one that has been placed before the court by the parties and that is “relevant to the performance of the judicial function and useful in the judicial process.” Amodeo I, 44 F.3d at 145. Next, if the record sought is determined to be a judicial document, the court proceeds to “determine the weight of the presumption of access” to that document. United States v. Erie Cty., 763 F.3d 235, 239, 241 (2d Cir. 2014). The weight to be accorded is “governed by the role of the material at issue in the exercise of Article III judicial power and the resultant value of such information to those monitoring the federal courts.” Amodeo II, 71 F.3d at 1049.6 Finally, the court must identify all of the factors that legitimately counsel against disclosure of the judicial document, and balance those factors against the weight properly accorded the presumption of access. Id. at 1050. Although the public’s right is strong, it is “not absolute.” Nixon, 435 U.S. at 598. Countervailing considerations that courts may consider include “the danger of impairing law enforcement or judicial efficiency” and “the privacy interests of those resisting disclosure.” Amodeo II, 71 F.3d at 1050. If, at the end of this process, the balance of the factors tips against permitting public access, then the court may deny disclosure. A. Judicial Documents7 The first inquiry — whether the video excerpts are judicial documents — is easily resolved here. Under our decisions in CBS and Graham, there can be little doubt that the video excerpts played for the jury constitute judicial documents.8 In CBS, we held “that the common law right to inspect and copy judicial records applies to videotaped depositions of witnesses,” and allowed the television network access to such videos for copying and potential broadcast. 828 F.2d at 959. As we explained in that decision, judicial documents include “any item entered into evidence at a public session of trial, excluding only those items entered under seal, but not distinguishing evidence on the basis of whether it was real or testimonial.” Id. (emphasis in original). That the parties did not seek admission as an exhibit either the entire video-recorded deposition at issue here or even just the portions shown to the jury makes no difference. Some other courts have distinguished between recordings that are introduced into evidence as exhibits and those that merely serve as a substitute for live testimony. See, e.g., Apple iPod iTunes Antitrust Litig., 75 F. Supp. 3d 1271, 1274 (N.D. Cal. 2014) (declining to apply a presumption of access to video of deposition that was played to the jury but was not admitted into evidence as exhibit). Our Circuit, however, rejected that distinction in Graham, where we held that video and audio recordings played at a pre-trial detention hearing but not introduced into evidence as exhibits were nonetheless judicial documents subject to the presumption of public access. 257 F.3d at 152. In so doing, we held that “the definition of a ‘judicial document’…extend[s] to any material presented in a public session of court relevant to the performance of the judicial function and useful in the judicial process whether or not it was formally admitted.” Id. at 153 (emphasis added). The District Court thus correctly ruled that those portions of the video-recorded deposition that were played for the jury in the Mirlis proceedings are judicial documents and are therefore subject to the common law presumptive right of access, notwithstanding that they were not formally introduced into evidence. B. Weight of Presumptive Right of Access Turning to the second step: The general and deeply rooted rule is that the presumptive right of access is afforded “strong weight” when applied to documents that play a central role in “determining litigants’ substantive rights — conduct at the heart of Article III.” Amodeo II, 71 F.3d at 1049. As this Court has observed many times, such access is critical as it enables the public to monitor the actions of the courts and juries to ensure “a measure of accountability” and bolster “confidence in the administration of justice.” See id. at 1048. See generally Globe Newspaper Co. v. Superior Court, 457 U.S. 596, 610 (1982) (“[A] presumption of openness inheres in the very nature of a criminal trial under our system of justice.”); Richmond Newspapers, Inc. v. Virginia, 448 U.S. 555, 573 (1980) (plurality opinion) (explaining that right of access “contributes to public understanding of the rule of law and to comprehension of the functioning of the entire criminal justice system”). The presumption of access is thus fundamental and not casually overcome. The excerpts of Hack’s video deposition that were played for the jury gave important, even crucial, corroboration for Mirlis’s account of sexual assault by Greer while Mirlis was a student at the Yeshiva. His demeanor in so testifying over the several hours played to the jury surely conveyed information that bore on the jury’s assessment of his account. That those excerpts played a central role in the jury’s determination of Mirlis’s and Greer’s substantive rights was implicitly confirmed by the jury’s request, which the District Court granted, to review portions of that video during their deliberations. Hack does not argue otherwise on appeal. The court thus correctly accorded a strong presumption of public access to the excerpts of Hack’s video deposition. C. Identification and Balancing of Countervailing Interests At the third step, however, we identify error in the District Court’s analysis. Foremost among the competing concerns that a court weighing disclosure must consider is “the privacy interest of the person resisting disclosure.” Amodeo II, 71 F.3d at 1050 (reiterating that “the privacy interests of innocent third parties…should weigh heavily in a court’s balancing equation”). Such interests establish a “venerable common law exception to the presumption of access.” Id. at 1051. In determining the proper weight to accord an asserted privacy interest, a court should consider both “the degree to which the subject matter is traditionally considered private rather than public,” as well as “[t]he nature and degree of the injury” to which the party resisting disclosure would be subjected were the privacy interest not protected. Id. The latter inquiry entails “consideration not only of the sensitivity of the information and the subject but also of how the person seeking access intends to use the information.” Id. As further described below, we conclude that the District Court erred by failing to address and accord weight to Dressler’s motives in seeking to obtain the video deposition and his intentions with respect to that video, and to the injury to Hack — as a minor victim of sexual abuse — that disclosure of the video would inflict, even after the substance of Hack’s testimony became public at trial. Although the court seemed to consider its decision granting Dressler access to be dictated by our 1987 decision in CBS, in this reading it erred: Unlike in CBS, Mirlis’s civil litigation did not involve a crime of national importance; the core information it conveyed was already public and had been publicized; the video recording at issue was of a highly sensitive and personal nature; and — perhaps most relevantly — the Internet’s rise over the last 30 years has had tremendous implications for the ease and immediacy of access to videos, as well as the permanence of those videos, increasing the potential for needless emotional harm to minor victims of sexual assault who seek to avoid being victimized even further. See, e.g., Paroline v. United States, 572 U.S. 434, 440 (2014) (discussing how the Internet can create a “permanent record” of abuse and “memorializ[e] the sexual assault and other sexual exploitation of children”). 1. Dressler’s Motives and Likely Use In January 2018, the District Court heard argument on Dressler’s request to release for copying the entire video of Hack’s deposition and certain other materials.9 In particular, this was a request to enable Dressler himself to copy the video, as no other person or entity had sought to do so. In his letter to the District Court seeking access, Dressler expressed condemnation for Greer’s acts of abuse and personal disdain for what he saw as Greer’s attempts to protect his (Greer’s) assets from collection of the $21.7 million judgment that was entered in Mirlis’s favor. Dressler further advised the court that Greer had filed, but not pursued, a defamation suit against Dressler. At the court’s hearing, Dressler affirmed his intention to post the video on his blog or in “some other place where [the video] would be available to anyone who has the internet.” App’x 449. During argument, Greer’s attorney asserted that Dressler was motivated to publicize the video by his desire to prejudice Greer in any future criminal proceedings. The District Court rejected Greer’s stated concern about Dressler’s motives as irrelevant, based on its understanding that our Court “has made clear that [Dressler's] motivations are beside the point in terms of the First Amendment and presumption of access issues.” Id. at 459. The court’s refusal to consider Dressler’s motives in obtaining the video and intentions for use of the video if he did obtain it rests on a mistaken overreading of our precedent. Although (as noted above) our Court has commented that the motive of a person seeking access to judicial documents is “irrelevant to defining the weight accorded the presumption of access,” Amodeo II, 71 F.3d at 1050 (emphasis added), we have never held that motive has no bearing on the broader task of balancing that presumption against considerations that counsel against disclosure. On the contrary, we have explained that courts should consider personal motives (including an applicant’s desire to pursue an “individual vendetta”) at the third, balancing step of the inquiry, in connection with any asserted privacy interests, “based on an anticipated injury as a result of disclosure.” Id. We have instructed courts that weighing “[t]he nature and degree of injury” requires “consideration not only of the sensitivity of the information and the subject but also of how the person seeking access intends to use the information.” Id. at 1051 (emphasis added). This information is relevant particularly when an individual, and not a news media organization, seeks copying access to sensitive information: As we cautioned in Amodeo II, personal vendettas “need not be aided” by the court. Id. Our reasoning in Amodeo II aligns closely with the Supreme Court’s observation in Nixon that courts may on occasion deny public access to judicial documents to ensure that judicial records do not “become a vehicle for improper purposes.” Nixon, 435 U.S. at 598. The Nixon Court expressed particular concern about protecting from purely prurient interest the display or disclosure of otherwise embarrassing private or familial information obtained through the courts. See id. at 601. It explained that “the common-law right of inspection,” although undoubtedly powerful and never lightly subordinated, must give way where records are sought merely “to gratify private spite or promote public scandal through the publication of the painful and sometimes disgusting details” of cases. Id. at 598; accord Amodeo II, 71 F.3d at 1051 (“Courts have long declined to allow public access simply to cater to a morbid craving for that which is sensational and impure.”). This concern is amplified where, as here, a video recording involves primarily conduct that is not of national or statewide importance and where the video is not of the criminal acts themselves, but of testimony regarding those acts. This case thus stands in sharp contrast, for example, to In re NBC, where we affirmed a district court’s decision to allow “three major television networks to copy and televise all videotapes admitted into evidence” in the nationally publicized ABSCAM criminal prosecutions, where those videos showed “some of the dealings between the undercover operatives and the four appellants, notably the acceptance by [one] Congressman…of $50,000 cash and his demand of an additional $35,000.” 635 F.2d 945, 947 (2d Cir. 1980). Thus believing itself bound to ignore them, the District Court disregarded the ample evidence of Dressler’s unsavory motives with respect to Greer for seeking access to Hack’s video testimony regarding Greer’s predatory intimacy with him as a young person. Dressler’s blog also evinces his impure motives with respect to Hack — a nonparty who, we note again, was primarily a victim of the defendant in, and not even a party to, the underlying litigation. True, Dressler’s letter to the court requesting access to, among other items, the video deposition expresses animus primarily toward Greer, not Hack.10 But to take that letter in isolation misses the forest for a tree: The blog on which Dressler announced that he planned to post Hack’s video deposition contains now, and contained at the time of the District Court’s decision, numerous postings — many of them published before the District Court’s hearing on Dressler’s request for public access to Hack’s video deposition. Those postings are dedicated to, and demonstrate considerable personal hostility toward, Hack, not just Greer.11 Dressler’s posts reveal at least his willingness, and as plausibly his desire, to publicize and try to draw attention to personal details of Hack’s life — including the location of Hack’s place of work and his home address — apparently with the intent to humiliate and harass him. His words and actions provide a legitimate basis for Hack’s concern that, by making a recording of Hack’s deposition available in whole or in part to Dressler, never mind to the public more generally, the District Court could be allowing its order to become “a vehicle for improper purposes.” Nixon, 435 U.S. at 598. We therefore conclude that the District Court erred in dismissing as irrelevant, at the balancing stage of its analysis, Dressler’s motivations, including those suggesting a probability that he was pursuing not the public interest, but personal vendettas. 2. Nature and Degree of Injury to Hack We further conclude that the District Court failed to give adequate weight to Hack’s privacy interests. At the threshold, the court set the bar for what “amounts to the sort of ‘extraordinary circumstance’ that would justify keeping it from the public” decidedly too high in the context presented. Hack App’x 143. In the District Court’s view, the subject of the deposition video — although it concerned in substantial part Hack’s experience, as a minor, of sexual abuse by a respected religious figure and elder — simply did not amount to “[t]he archetypical extraordinary case,” which was “far more gruesome.” Id. The court provided as a contrasting example “videos made by [a] kidnapper of [a] blindfolded and bound rape victim,” such as a district court denied media access to in a 1980 case in another jurisdiction. Id. (citing In re KSTP Television, 504 F. Supp. 360 (D. Minn. 1980)). KSTP, the decision that the District Court pointed to, involved a request by a commercial television station “to view and copy some three hours of tapes received in evidence in a criminal case.” 504 F. Supp. at 361. The videos that the network sought depicted the criminal defendant, who by then had been convicted of kidnapping, interacting with one of his victims. Id. Of the nine hours of footage at issue, the court had received into evidence three hours. Id. Those hours contained images of “conversations and conduct preliminary to, and anticipatory of, the actual sexual acts” the defendant committed on his victim. Id. The KSTP court denied the network’s request, resting its decision on the nature of the conduct, its observation that “[a]ll of the information in the tapes has already been made available to the public,” and the high value it assigned the innocence and privacy of the victim. Id. at 362-64 (concluding that there was “no public interest to be served by release of the tapes” here). Some years later, in 1987, we decided CBS. There, we reversed a district court’s denial of the television network’s application to copy for possible broadcast a witness’s video deposition that had been shown in full in open court during a criminal trial. The deponent in CBS was an imprisoned former union leader who “appeared ill in the videotape and was compelled to testify concerning his involvement in illegal activities.” CBS, 828 F.2d at 961. In ordering the public release of the video deposition, we distinguished the case from KSTP, citing the videos that were at issue in KSTP as an example of judicial documents to which public access could lawfully be precluded. See id. But our citation there of KSTP should by no means be read to establish a hurdle of “gruesomeness” that other cases must clear to overcome the common law presumption of access, powerful as it is; indeed, the reasoning of the KSTP court more resonates with ours here. Nor is CBS determinative of the outcome here, as our primary focus in CBS was whether the presumption of public access applied at all to video recordings of depositions. We devoted very little discussion to the opposing party’s countervailing privacy interests, of which there were but few: In granting access in CBS, we observed only that “[o]ld age and ill health are neither uncommon nor generally a cause of severe embarrassment,” and concluded that the situation of the elderly and ill deponent, who was confessing to his own criminal conduct, “is simply not analogous to, say, that of a victim of a slasher.” Id. Other circumstances present in CBS also make that decision less apposite here: While the CBS deponent had been convicted of crimes relevant to his testimony, Hack has not been convicted of, or even charged with, any crime. While the CBS deponent described his own criminal activity, Hack described being a victim of a crime when he was a minor. And while the CBS deponent’s advanced age and illness were neither uncommon nor particularly embarrassing, Hack’s description — of a sexual relationship that a trusted religious leader over 30 years his senior perpetrated in Hack’s youth and into his young adulthood — is likely both. Finally, that the video at issue here also contains Hack’s admission to his own wrongful behavior — namely, his failure to report Greer’s abuse of Mirlis — seems of relatively little import in the context of Mirlis’s suit against Greer. In sum, Hack is more akin to the “innocent third part[y]” whose privacy interests “ weigh heavily in a court’s balancing equation,” Amodeo II, 71 F.3d at 1050, than he is to the CBS deponent whose circumstances were “solely the result of his criminal acts,” CBS, 828 F.2d at 961. In addition, although the District Court acknowledged that courts are generally protective of the privacy interest of non-parties like Hack, it discounted the weight due that interest in Hack’s case. It gave three reasons: (1) Hack’s testimony had already been made publicly available in transcript form; (2) Hack effectively consented to the publication of his deposition testimony when he agreed to sit for a deposition without requesting authority to use a pseudonym or that the deposition be sealed; and (3) Hack evaded process and refused to appear and give live testimony at trial, a decision that created the need to present his video testimony to the jury in the first place. Each one of these rationales for discounting Hack’s privacy interests is flawed. With respect to the first rationale, the availability of a transcript of the deposition does not in our view necessarily eliminate or even diminish a party’s privacy interest in the publication or copying of a video of those proceedings. To the contrary: That the substance of the desired content is publicly available in some format (i.e., a transcript) tends in the circumstances presented here to cut against the public interest in the release of the content in a different form (i.e., video), since the primary public interest — general availability of the relevant information — has already been served.12 We break no new ground by articulating this principle. The Supreme Court in Nixon explained, in declining to compel the release of audio tapes, that the public interest in accessing audio recordings is weaker where “[r]eporters also were furnished transcripts of…tapes,” reasoning that the fact that “[t]he contents of th[ose] tapes were given wide publicity” negated any “question of a truncated flow of information to the public.” 435 U.S. at 609. Here, as in Nixon, the substantive information conveyed to the jury in the video of Hack’s deposition has been made public and has been written about in the local press, to all appearances largely satisfying the legitimate public interest in the trial. What remains private is solely the video recorded images of Hack actually saying these words — the publication of which, especially on the Internet, would impose a significant burden on Hack by immediately and forever connecting the extremely personal content of his testimony with his likeness, exposing his emotions as a victim. We have not set an absolute rule that the public availability of a deposition transcript guarantees the court’s protection of a deposition video, nor do we do so now. As we observed in CBS, “[v]ideotaped depositions…convey the meaning of testimony more accurately and preserve demeanor evidence as well.” 828 F.2d at 960. These undoubtedly are valuable components of the truth-finding process. The general rule of production that we applied in 1987 in CBS thus remains vital today. But we must also acknowledge what has changed since we decided CBS in 1987: The astonishing and pervasive rise of the Internet; the attendant ease with which videos may be shared worldwide by individuals; and the eternal digital life with which those videos are likely endowed by even a single display online. These are all factors that multiply and intensify the privacy costs to the individual of releasing sensitive videos; those costs are undeniably greater than what they might have been 30 years ago. Whereas the subject of a video deposition made public in 1987 may have suffered brief notoriety and embarrassment as the subject of an evening’s newscast, today, Hack could reasonably fear that, for the rest of his life, this video would be the first result of an internet search for his name. Given the proliferation of smartphones and improved digital streaming capabilities, he could also reasonably expect, as a schoolteacher and father, that his students and his children would view the video not only at home, on family computers, but possibly also during (his) class, on their cell phones. Common sense and over two decades of widespread and constant use of the Internet are sufficient to tell us that a video of a person describing details of his abuse is likely to garner more attention, be distributed more widely, and last longer in the public’s attention than are copies of a transcript or even local news articles. With respect to the second rationale, Dressler argues that we should construe Hack’s failure to request a pseudonym for use during the deposition, to seek restrictions on the video’s airing to the jury, or to move to seal the transcript, as amounting to his consent to general publication of the entire video deposition. But this can hardly be a fair conclusion. As explained above, the dissemination of the deposition video and the publication of the transcript impose very different privacy burdens, regardless of the fact that the transcript uses Hack’s name. Hack’s failure to request a pseudonym or the sealing of the transcript thus has little bearing on his privacy interest in the video, nor would it have had the effect of disguising his identity, given his unique position at the Yeshiva during the years in question. For the same reason, Hack’s failure to seek to prevent the video deposition from being played at trial — where it would be shown likely just once, to a limited audience, and in a venue where electronic recording is generally prohibited — cannot reasonably be treated as implicit consent to the video’s wider publication across the Internet. In certain instances, a person who declines to take reasonable steps to protect his or her private information may reasonably be understood to stake a lesser privacy claim in that information. Here, however, Hack diligently and timely objected to Dressler’s application. Only rarely will voluntary provision of sensitive evidence reasonably be understood to constitute consent to the widespread and likely permanent dissemination of a visual digital record of the formal encounter through which that evidence was given, when the encounter itself is not the allegedly critical act. With respect to the third rationale, Hack’s evasion of service, while certainly wrongful, was an attempt to protect his privacy by avoiding testifying in open court about sensitive and embarrassing subjects. Whatever sanction a non-party witness might merit for such actions, publicly releasing the very sensitive images that the witness so fiercely struggled to keep private, where such release may not otherwise be warranted, is not among them, contrary to Dressler’s argument. A holding otherwise could be expected to disincentivize naturally reluctant victim-witnesses from facilitating their depositions in the first place; it would hardly encourage them to give live testimony at trial. * * * In sum, we agree that the public is due a strong presumptive right of access to the portion of the video of Hack’s deposition that was shown at trial, and that right is not easily overcome, even where a transcript is readily available and the public has already had wide access to the substance of Hack’s testimony. But Dressler’s motives in obtaining, and intent in releasing, the video erode against the weight due that presumption here. So, too, do Hack’s countervailing privacy interests as a minor victim of sexual abuse perpetrated by a trusted religious figure in the community. Our 1987 decision in CBS does not dictate the outcome of the balancing of concerns that should apply in Hack’s case in 2020. The current record allows us to balance the relevant considerations properly. Having done so, we REVERSE the District Court’s order and REMAND the cause with instructions that the District Court enter an order denying Dressler’s request for the video recording. CONCLUSION For the foregoing reasons, the District Court’s order granting Dressler’s request for production is REVERSED.
