Opinion for the Court filed by Circuit Judge MILLETT.
Removing the cloak from the cloak-and-dagger business of spying can be a lengthy and arduous process. Understandably so, given the competing needs to protect national security and to ensure appropriate governmental transparency. The 30-year odyssey of this Freedom of Information Act ease attests to the complex twists and turns that the disclosure process can take.
In 1985, Carl Oglesby filed a request under the Freedom of Information Act with six federal agencies, seeking information on Reinhard Gehlen, a former Nazi general through whom the United States engaged in clandestine espionage after World War II. Thirty years, an intervening Act of Congress, and two appeals later, more than ten thousand pages of documents have been released and the quest for information has narrowed substantially. With Mr. Oglesby’s passing in 2011, his daughter, Aron DiBacco, and partner, Barbara Webster, have now taken up Oglesby’s cause. In this third appeal, DiBacco and Webster challenge the adequacy of the Army’s and CIA’s searches for and *183disclosures of documents, as well as the CIA’s justification for withholding certain information on national security grounds.
The district court concluded that the Army and CIA have done what the Freedom of Information Act requires.. We agree, except that we must remand for the district court to address in the first instance DiBacco’s and Webster’s challenges to redactions in a batch of records that the Army disclosed to them while this appeal was pending.
I
Statutory Framework
The Freedom of Information Act
Congress enacted the Freedom of Information Act (“FOIA”), 5 U.S.C. § 552, to promote the “broad disclosure of Government records” by generally requiring federal agencies to make their records available to the public on request. Department of Justice v. Julian, 486 U.S. 1, 8, 108 S.Ct. 1606, 100 L.Ed.2d 1 (1988) (internal quotation marks omitted). But Congress also “realized that legitimate governmental and private interests could be harmed by release of certain types of information.” Id. (internal quotation marks omitted). Accordingly, FOIA “balance[s] the public’s need for access to official information with the Government’s need for confidentiality,” Weinberger v. Catholic Action of Hawaii, 454 U.S. 139, 144, 102 S.Ct. 197, 70 L.Ed.2d 298 (1981), by exempting nine categories of records from disclosure, see 5 U.S.C. § 552(b). While those exemptions “must be narrowly construed,” Milner v. Department of Navy, 562 U.S. 562, 565, 131 S.Ct. 1259, 179 L.Ed.2d 268 (2011) (internal quotation marks omitted), they still must be given “meaningful reach and application,” John Doe Agency v. John Doe Corp., 493 U.S. 146, 152, 110 S.Ct. 471, 107 L.Ed.2d 462 (1989).
FOIA Exemptions 1 and 3 are at issue in this case. Exemption 1 authorizes the withholding of “matters” that are “specifically authorized under criteria established by an Executive order to be kept secret in the interest of national defense or foreign policy” if they “are in fact properly classified pursuant to such Executive order.” 5 U.S.C. § 552(b)(1).
Exemption 3 excludes “matters” that are “specifically exempted from disclosure by statute” if that statute “requires that the matters be withheld from the public in such a manner as to leave no discretion on the issue” or “establishes particular criteria for withholding or refers to particular types of matters to be withheld!.]” 5 U.S.C. § 552(b)(3). Courts have held that a provision of the National Security Act of 1947, which calls for the Director of National Intelligence to protect “intelligence sources and methods from unauthorized disclosure,” 50 U.S.C. § 3024(i)(1), is a valid Exemption 3 statute. CIA v. Sims, 471 U.S. 159, 167, 105 S.Ct. 1881, 85 L.Ed.2d 173 (1985); accord Larson v. Department of State, 565 F.3d 857, 865 (D.C.Cir.2009).
Under FOIA, agencies may charge reasonable fees to help defray their costs in responding to a FOIA request, but they must waive or reduce their fees if disclosure of the requested information “is in the public interest because it is likely to contribute significantly to public understanding of the operations or activities of the government and is not primarily in the commercial interest of the requester.” 5 U.S.C. § 552(a)(4)(A)(iii).
When an agency subject to FOIA receives a request for records, it must determine within twenty days whether to comply with that request and, once it does, must immediately notify the requester of its determination and reasoning. 5 U.S.C. *184§ 552(a)(6)(A)(i). Upon receipt of that determination, the requester may administratively appeal the agency’s decision, and the agency must decide the appeal within twenty- days. See id. § 552(a)(6)(A)(ii). Exhaustion of that administrative appeal process is a prerequisite to seeking judicial relief, unless the agency has not responded within the statutory time limits. See id. § 552(a)(6)(C); Oglesby v. Department of Army (Oglesby I), 920 F.2d 57, 61-62 (D.C.Cir.1990).
Federal district courts have jurisdiction under FOIA “to enjoin [an] agency from withholding agency records and to order the production of any agency records improperly withheld from the complainant.” 5 U.S.C. § 552(a)(4)(B). In a FOIA suit, the burden is “on the agency to sustain its action,” and the district court must “determine the matter de novo.” Id.
The Nazi War Crimes Disclosure Act
Congress enacted the Nazi War Crimes Disclosure Act (“Disclosure Act”), Pub.L. No. 105-246, 112 Stat. 1859 (1998) (codified as amended at 5 U.S.C. § 552 note), to spur disclosure of millions of pages of government records from the World War II era. See Nazi War Crimes & Japanese Imperial Government Records Inter-agency Working Group, Final Report to the United States Congress 1 (April 2007) (“Interagency Report”). To that end, the Disclosure Act required federal agencies to “locate, identify, inventory, recommend for declassification, and make available to the public at the National Archives and Records Administration,” with few exceptions, any remaining classified records concerning war crimes committed by Nazi Germany and its allies. Pub.L. No. 105-246, § 2(c)(1).
The Disclosure Act also directed the President to establish the “Nazi War Criminal Records Interagency Working Group,” Pub.L. No. 105-246, § 2(b)(1), composed of various high-level government officials and three members of the public. The Working Group was tasked with coordinating agencies’ efforts to fulfill the Disclosure Act’s mandate. See Interagency Report, at l.1 Those efforts led to the declassification and public release of over 8.5 million pages of World War II and post-war records. See Interagency Report, at 2.
Factual and Procedural Background
General Reinhard Gehlen served as Hitler’s senior military intelligence officer on the Eastern Front. See Oglesby I, 920 F.2d at 60. After the war, Gehlen became an intelligence asset for the United States, secretly agreeing to operate an extensive spy network in Europe under United States command. See id. at 60. Gehlen operated this spy network, known as the Gehlen Organization, until 1956, at which point it became part of the newly formed intelligence service of the Federal Republic of. Germany. Gehlen led the latter until his retirement in 1968. Interagency Report, at 11, 13, 30, 48; CIA Biographic Sketch on General Reinhard Gehlen, NWC-002652 (declassified and approved *185for release under the Disclosure Act in 2001), J.A. 1084-1085.
Carl Oglesby was a journalist interested in the intelligence relationship between the United States and Gehlen. His rounds of effort over many years to obtain information under FOIA contributed materially to the disclosure of the Gehlen Organization’s covert relationship with the federal government.
Round One
In 1985, Oglesby submitted FOIA requests seeking information on that relationship to six federal agencies: the Department of the Army, the Department of State, the Federal Bureau of Investigation, the Central Intelligence Agency, the National Archives and Records Administration, and the National Security Agency (“NSA”). See DiBacco v. Department of Army, 983 F.Supp.2d 44, 49 (D.D.C.2013). Oglesby specifically sought:
(i) Records of World War II German General Reinhard Gehlen and his relationship with any United States officials during the period 1944 through 1956;
(ii) Records of the meetings held at Fort Hunt, Virginia, in the summer of 1945 between Gehlen and American officials including U.S. Army General George V. Strong and Office of Strategic Services officer Allen Welsh Dulles;
(in) Records of U.S. Army “Operation Rusty,” carried out in Europe between 1945 and 1948;
(iv) Records of post-war Nazi German underground organizations such as Odessa, Kamaradenwerk, Bruders-chaft, Werewolves, and Die Spanne; and
(v) Records of the Office of Strategic Services’ “Operation Sunrise” in 1945.
Complaint ¶¶ 5, 23, 34, 40, 57, 63; J.A. 54 (request to CIA); J.A. 79 (request to FBI); J.A. 97 (request to National Archives); see also Oglesby v. Department of Army (Oglesby II), 79 F.3d 1172, 1175-1176 (D.C.Cir.1996). Oglesby also sought a waiver of search and copying fees from each agency under 5 U.S.C. § 552(a)(4)(A). Oglesby I, 920 F.2d at 60.
The agencies collectively released 384 pages of documents (many redacted) in response, invoking various FOIA exemptions as a basis for refusing further disclosures. See Oglesby I, 920 F.2d at 66-71. The Army, CIA, and National Archives also rejected Oglesby’s fee-waiver requests. The NSA in due course agreed to waive its fees. Id.
In 1987, Oglesby filed suit under FOIA in the United States District Court for the District of Columbia, arguing that the agencies had performed inadequate searches for responsive documents, failed to properly support their exemption claims, and wrongly refused to waive their fees. Oglesby I, 920 F.2d at 61. The district court granted summary judgment to the agencies on all issues. Id.
This court reversed. We first held that only claims against one agency — the State Department — were before us, because Oglesby had failed to exhaust his claims with the other five agencies. We further held that the State Department had provided insufficient details about its search for documents to support summary judgment. Oglesby I, 920 F.2d at 71. We vacated and remanded for further proceedings on the adequacy of the State Department’s search and for Oglesby to exhaust his claims with the other agencies. Id. We did not reach Oglesby’s fee-waiver arguments, but suggested that the Army reconsider its denial of a waiver in light of the NSA’s decision to grant one. Oglesby I, 920 F.2d at 66 n. 11.
*186 Round Two
Several years later, having exhausted his administrative remedies with the other five agencies without satisfactory resolution, Oglesby returned to district court. See Oglesby II, 79 F.3d at 1175. Shortly after Oglesby filed his complaint seeking further disclosures by all six agencies, the Army and CIA granted Oglesby fee waivers, and the Army, CIA, and NSA released additional documents. See id. at 1179-1184. All six agencies filed affidavits describing their searches for documents, and those agencies that had withheld documents included Vaughn indices to justify their withholdings. See id. at 1176.2 The district court again granted summary judgment for all six agencies. Id.
We, again, reversed in part. Oglesby II, 79 F.3d at 1175. We held that the Army, CIA, and NSA had failed to adequately justify their withholdings in their Vaughn indices, and that the Army and CIA had failed to establish the adequacy of their searches. Id. With respect to all claims against the State Department, FBI, and National Archives, we affirmed the district court’s grant of summary judgment. Id. Round Three
While the case against the Army, CIA, and NSA was pending on remand, intervening legislation — the 1998 Disclosure Act — significantly altered the legal landscape. That Act led the CIA to revisit its stance toward records concerning General Gehlen. Previously, because the CIA’s relationship with Gehlen had been classified, the agency issued so-called Glo-mar responses “neither confirming] nor denying] the existence or nonexistence of responsive records” to Oglesby’s requests. McNair Decl. ¶¶ 2-4 (Sept. 20, 2000), J.A. 594-596.3
In the wake of the Disclosure Act, the CIA filed a declaration in this case publicly acknowledging its relationship with General Gehlen for the first time. See McNair Decl. ¶¶ 4-10. The declaration explained that, even though the United States government did not consider General Gehlen to be a Nazi war criminal, the agency’s Disclosure Act searches uncovered other Nazi war criminal records that revealed the United States’ intelligence relationship with Gehlen. See id. ¶ 7; Interagency Report, at 48. Rather than withhold such information as beyond the purview of the Disclosure Act, the agency decided to declassify it.
That declassification, the CIA explained, would “have a significant impact upon this *187case” because the agency could now process Oglesby’s FOIA request on Gehlen. McNair Decl. ¶¶ 2, 10. A series of status reports followed in which the CIA, NSA, and Army laid bare the time-consuming task in front of them. In one early report, the CIA advised that recent searches had turned up “approximately 251 boxes of material, and 2,901 folders, with documents that likely contain records regarding General Gehlen.” .Defendants’ Status Report 1-2 (Dec. 11, 2000), J.A. 622-623.
The CIA proposed consolidating its response to Oglesby’s FOIA request with its ongoing efforts under the Disclosure Act. Defendants’ Status Report at 2. Estimating that it would take a year to complete its Disclosure Act process, the CIA asked for two years to finish action on Oglesby’s FOIA request because it anticipated finding and processing “additional documents that go beyond the scope of the [Disclosure] Act[.]” Id. The Army, for its part, joined the CIA’s request, advising that it, too, would need to review additional materials. Id. at 2-3.
During a status hearing in January 2001, agency counsel again indicated that the CIA “now has 251 boxes,” each “one cubic foot in size” and expected to be “full” “of material regarding the Gehlen organization,” totaling “anywhere between 251,-000 and 775,000 pages[.]” Tr. of Hearing 4:15-23 (Jan. 9, 2001), J.A. 630.
In a subsequent status report, the agency explained that, through extensive search efforts, it had “identified [a] potential universe of over 25,000 responsive documents.” Defendants’ Status Report 2 (Feb. 5, 2001), J.A. 648. After reviewing those documents, the CIA would forward them to the National Archives for public release. Id. Once the National Archives provided the CIA with a final release copy of the documents, the CIA would provide Oglesby with any documents responsive to his FOIA request. Id. at 3. The CIA reiterated its two-year estimate for the project’s completion. Id. at 4.
After that flurry of activity, the case languished for over a decade, with no substantive action by the court, the agencies, or Oglesby. Oglesby passed away in September 2011.
Round Four
In December 2011, Oglesby’s daughter, Aron DiBacco, and domestic partner, Barbara Webster (collectively, “DiBacco”), filed a motion to substitute themselves as plaintiffs, which the district court granted. DiBacco, 983 F.Supp.2d at 52. Both the agencies and DiBacco then filed for summary judgment. DiBacco also moved to compel disclosure of classified declarations referenced in the agencies’ summary judgment briefing the previous decade.
The district court granted summary judgment for the agencies. As a preliminary matter, the court ruled that the Interagency Report (prepared by the Disclosure Act’s Interagency Working Group to document its efforts under the Act) was a “record” of a “public office” admissible under Federal Rule of Evidence 803(8). DiBacco, 983 F.Supp.2d at 54-55. The court also declined to compel disclosure of the previously filed classified declarations because it had neither reviewed nor relied on them in any proceeding. Id. at 53.
The court next ruled that the CIA had conducted an adequate search for documents. DiBacco, 983 F.Supp.2d at 55-58. Relying .on the Interagency Report and on the CIA’s declarations, the court determined that the CIA’s Disclosure Act searches were reasonably calculated to locate records responsive to Oglesby’s FOIA request. Id. at 56.
The court also upheld the CIA’s with-holdings under Exemptions 1 and 3. DiBacco, 983 F.Supp.2d at 58-61. With re*188gard to Exemption 1, the court concluded that, contrary to DiBacco’s claim, certain documents were properly classified decades ago, with -markings required under the then-governing Executive Orders. Id. As for Exemption 3, the court concluded that the CIA retained the power to withhold documents to protect intelligence sources and methods under the National Security Act. Id. at 61.
Turning to the Army, the district court ruled that its search for responsive records was proper. DiBacco, 983 F.Supp.2d at 61-64. While DiBacco argued that the Army had transferred responsive documents to the National Archives to evade its FOIA obligations, the district court ruled that, in fact, the transfer facilitated public access to the documents as required by the Disclosure Act. Id. at 63-64. The district court also rejected DiBacco’s challenge to the Army’s search, relying again on the Interagency Report and declarations describing the search. Id. at 62-64.
As for the NSA, the district court ruled that it had adequately justified its withholding of records. DiBacco, 983 F.Supp.2d at 64-66. DiBacco has not challenged that ruling.
II
Analysis
The district court had jurisdiction under 28 U.S.C. § 1331 and 5 U.S.C. § 552(a)(4)(B). We have jurisdiction oVer the district court’s final judgment under 28 U.S.C. § 1291.
Claims Against the Army
DiBacco contends that the Army failed to conduct an adequate search for documents. She also maintains that the Army violated FOIA by transferring relevant documents to the National Archives. Neither argument succeeds.
The Army’s Search for Records
We review de novo the adequacy of the Army’s search. See Valencia-Lucena v. United States Coast Guard, 180 F.3d 321, 326 (D.C.Cir.1999). The burden is on the agency to demonstrate that it made a “good faith effort to conduct a search * * * using methods which can be reasonably expected to produce the information requested.” Oglesby I, 920 F.2d at 68. Courts may rely on a “reasonably detañed affidavit, setting forth the search terms and the type of search performed, and averring that all files likely to contain responsive materials (if such records exist) were searched.” Valencia-Lucena, 180 F.3d at 326 (internal quotation marks omitted). Summary judgment must be denied “if a review of the record raises substantial doubt, particularly in view of well defined requests and positive indications of overlooked materials[J” Id. (internal quotation marks omitted).
The Army relies mainly on its search efforts under the Disclosure Act to demonstrate the adequacy of its search for documents responsive to Oglesby’s FOIA request. Much of that process is described in the Interagency Report that the Inter-agency Working Group submitted to Congress. The district court relied on that report in granting summary judgment. DiBacco, 983 F.Supp.2d at 54-55. The court also relied on the Army’s declarations — one from Martha Wagner Murphy, Chief of the Special Access and FOIA Branch at the National Archives, and two from Bradley Dorris, Director of the FOIA and Investigative Records Office at the United States Army Intelligence and Security Command (“INSCOM”). See Murphy Deck (Dec. 14, 2012), J.A. 689-700; First Dorris Déel. (Nov. 27, 2012), J.A. 701-702; Second Dorris Decl. (March 20, 2013), J.A. 1050-1052.
*189According to DiBacco, the district court should not have relied on the Interagency-Report because it was inadmissible. Not so. Federal Rule of Evidence 803(8), an exception to the hearsay bar, allows as evidence a “record or statement of a public office” where, as relevant here, (i) the document “sets out the office’s activities,” and (ii) “neither the source of information nor other circumstances indicate a lack of trustworthiness.”
The Interagency Report easily fits that bill: It is an official document prepared by the Interagency Working Group that sets out the Group’s activities, as statutorily required by the Disclosure Act. See Pub.L. 105-246, § 2(c)(3) (requiring the Interagency Working Group to “submit a report to Congress” describing all Nazi war. criminal records that it found, “the disposition of such records, and the activities of the [Interagency Working Group] and agencies under this section”).
DiBacco, in fact, does not dispute that the Interagency Report is a “record” of a “public office” that sets out “the office’s activities” within the meaning of Rule 803(8). She argues only that the Inter-agency Report is untrustworthy. But all that she points to is a statement in the report indicating that the Interagency Working Group “did not seek unanimous agreement on a single ‘official’ version of the[ ] declassification effort.” Interagency Report, at v.
DiBacco needs to read on. That statement goes on to say that that any personal or institutional perspectives from Group members would be included in a separate chapter. Interagency Report, at v. Nothing in that chapter casts any reasonable doubt on the Report’s account of the agencies’ search efforts. See DiBacco, 983 F.Supp.2d at 54; Interagency Report, at 81-101.
The Army’s search effort focused on classified intelligence and counterintelligence records maintained by INSCOM. See Interagency Report, at 52. Those files generally concerned (i) “foreign personnel and organizations,” (ii) “intelligence and counterintelligence sources,” and (iii) “counterintelligence security investigations.” Id. The records consisted of 13,000 reels of 35mm microfilm (holding approximately 1.3 million files), and approximately 460,000 individual paper files. Id. The Army created digitized images of the microfilm files, which it then searched electronically for responsive information using a database containing the names of Nazi officers and other individuals connected to Nazi war crimes. Id. at 29, 54. After reviewing and declassifying the relevant files under the Disclosure Act, the Army turned them over to the National Archives. Id. at 54. The Army also conducted a manual review of its remaining paper files.
Between 2000 and 2001, the Army transferred over 20,000 digitized and paper files to the National Archives. Interagency Report, at 54. The vast majority of those files were fully declassified, although some contained limited redactions. Id. The Army undertook further searches using additional relevant terms discovered by the Inter-agency Working Group and participating agencies. Id. The Army eventually transferred, in 2005, the original 13,000 reels of microfilm and a full set of about 1.3 million scanned microfilm files to the National Archives. Murphy Decl. ¶ 12(d); Inter-agency Report, at 54. The Army did not retain any copies of those files. Second Dorris Decl. ¶ 5.
Upon receipt, the National Archives “accepted] full responsibility for administering the files both technically and for reference purposes.” Murphy Decl. ¶ 11. As a consequence, the National Archives, rather than the Army, conducted the most recent *190searches of those files for .records responsive to Oglesby’s FOIA request. In so doing, the National Archives used a variety of keywords to locate records concerning or related to General Gehlen. See Murphy Decl. ¶¶ 13, 15-16. Also included within the scope of the search were common misspellings of various codenames, along with pseudonyms and codewords the CIA had separately created for Gehlen and the Gehlen Organization. See id. ¶¶ 15-16. The search yielded no records regarding meetings held at Fort Hunt in the summer of 1945 between General Gehlen and high-level United States officials, including George Strong or Allen Dulles. But the search did locate 2,863 pages of records responsive to other aspects of Oglesby’s FOIA request, all but 11 pages of which were fully declassified and have since been provided to DiBacco. Murphy Decl. ¶ 17-18; Letter from Vincent H. Cohen, Jr., Acting United States Attorney for the District of Columbia (April 1, 2015); Letter from James H. Lesar, Counsel for Appellants (April 8, 2015).
The Army’s declarations from Bradley Dorris at INSCOM confirm that the “records most likely responsive to the FOIA requests would have been in the Investigative Records Repository at INSCOM[,]” which had been transferred to the National Archives. First Dorris Deck ¶ 6. To be certain, INSCOM had conducted “an exhaustive search” of its hard copy and electronic files, which turned up nothing. Id. ¶ 5. In another declaration, Dorris clarified that “the records which would be responsive to the FOIA requests would have been in the Investigative Records Repository at INSCOM,” and that he was “unaware of any other locations of any records related to [Oglesby’s] FOIA request.” Second Dorris Deck ¶7. The requested documents, Dorris explained, “were intelligence files,” and so “the only location the documents would be located would be at INSCOM,” the Army’s sole “intelligence records repository.” Id.
DiBacco levels five challenges to the Army’s search, which did not produce certain materials she believes exist and had hoped to find. But FOIA is not a wishing well; it only requires a reasonable search for records an agency actually has.
First, relying on the initial Dorris declaration, DiBacco asserts that the Army improperly searched only the locations “most likely” to contain responsive documents, while FOIA requires it to search all locations “likely” to contain such documents. DiBacco is correct that “most likely” is not the relevant metric. See Oglesby I, 920 F.2d at 68 (“[T]he agency cannot limit its search to only one record system if there are others that are likely to turn up the information requested.”). But the point gains her nothing because Dorris clarified that the only place containing records “responsive to the FOIA requests would have been in the Investigative Records Repository at INSCOM,” and that he knew of no other locations that might contain responsive records. Second Dorris Deck ¶ 7. That declaration attests that the Army applied the proper search standard.
Second, DiBacco argues that the Army’s failure to turn up documents on secret meetings at Fort Hunt — documents that DiBacco feels certain must exist — demonstrates the inadequacy of the Army’s search. We put that losing claim to bed twenty-five years ago, see Oglesby I, 920 F.2d at 67 n. 13, and age has not improved it. Oglesby’s (now DiBacco’s) “conviction that the Fort Hunt meeting was of such importance that records must have been-created is pure speculation,” and “[s]uch hypothetical assertions are insufficient to raise a material question of fact with respect to the adequacy of the agency’s search.” Id.
*191DiBacco maintains that this time is different because the CIA has disclosed the Gehlen relationship and because “new evidence has emerged.” Appellants’ Br. 37. Those are not differences of any consequence. That the CIA has now acknowledged the Gehlen relationship does nothing to show that meetings at Fort Hunt ever took place. Moreover, DiBacco’s “new evidence” — two Washington Post articles — establishes only that prisoners were held and interrogated at Fort Hunt. The articles do not even hint at secret meetings. Absent a more substantial showing, the Army’s “failure to turn up a particular document, or mere speculation that as yet uncovered documents might exist, does not undermine the determination that the agency conducted an adequate search for the requested records.” Wilbur v. CIA, 355 F.3d 675, 678 (D.C.Cir.2004) (per curiam).
Third, DiBacco assails the adequacy of the Army’s search for records conducted before the CIA disclosed its relationship with Gehlen. That argument is moot, long since overtaken by the comprehensive searches undertaken under the Disclosure Act. Those searches have looked further and wider than FOIA requires. The declarations from the Army and the National Archives describe searches of Army records reasonably calculated to discover all documents responsive to Oglesby’s request. That additional Army documents were found at the National Archives through those efforts further substantiates the search’s adequacy. And adequacy— not perfection — is the standard that FOIA sets. See Oglesby I, 920 F.2d at 68. Beyond that, DiBacco provides no reason why the Army’s decades-old search would be germane to any remaining material issue, so we need not address that question.
Fourth, DiBacco cites a book on General Gehlen, written by Mary Ellen Reese, in which the author claimed to have filed FOIA requests yielding “well over a thousand documents” about Gehlen. Appellants’ Br. 48. That figure, DiBacco urges, far exceeds the number of documents the Army disclosed in response to Oglesby’s FOIA request.
Oglesby made that precise argument in Oglesby II, and we held that further explication of the Army’s search was needed. 79 F.3d at 1185. But a lot has changed in the intervening nineteen years. Most relevantly, following Oglesby’s success in the second appeal, the Army released 9,000 additional pages of responsive material to Oglesby, including thousands of pages related to Gehlen. And the Army has since provided 2,863 additional pages of responsive documents to DiBacco. DiBacco, for her part, has provided no further information on Reese’s request — such as its scope or the number of pages received — or any other basis for concluding that the Army is holding back documents.
Fifth and finally, DiBacco suggests two additional search terms that, in her view, the Army should have used: “GO,” an abbreviation for the “Gehlen Organization,” and “PO Box 1142,” a codename for Fort Hunt. Appellants’ Br. 48-49. But it is undisputed that the agencies searched for records pertaining to the Gehlen Organization and employed relevant code-names. The Army’s burden was to show that its search efforts were reasonable and logically organized to uncover relevant documents; it need not knock down every search design advanced by every requester. See SafeCard Services, Inc. v. SEC, 926 F.2d 1197, 1201 (D.C.Cir.1991). (“When a plaintiff questions the adequacy of the search an agency made in order to satisfy its FOIA request, the factual question it raises is whether the search was reasonably calculated to discover the requested *192documents, not whether it actually uncovered every document extant.”).
The Army’s Transfer of Documents
DiBacco contends that we cannot affirm summary judgment for the Army, no matter how thorough its search, because its transfer of documents to the National Archives under the Disclosure Act casts doubt on its motives. In DiBacco’s view, there is a genuine dispute over whether the Army transferred documents to avoid disclosing them to Oglesby. That argument beggars belief.
FOIA generally obligates covered agencies to disclose their records, unless they are exempted. But “possession or control is a prerequisite to FOIA disclosure duties[.]” Kissinger v. Reporters Committee for Freedom of the Press, 445 U.S. 136, 152, 100 S.Ct. 960, 63 L.Ed.2d 267 (1980). Accordingly, when an agency does not possess or control the records a requester seeks, the agency’s non-disclosure does not violate FOIA because it has not “withheld” anything. Id. at 150, 100 S.Ct. 960. “[A]n agency has no duty to retrieve and release documents it once possessed but that it legitimately disposed of prior to the date a FOIA request was received.” Chambers v. Department of Interior, 568 F.3d 998, 1004 (D.C.Cir.2009) (quoting McGehee v. CIA, 697 F.2d 1095, 1103 n. 33 (D.C.Cir.1983)).
This ease, though, is a bit more complicated because the Army transferred its documents to the National Archives after receiving Oglesby’s FOIA request. The general rule is that an agency may not avoid a FOIA request by intentionally ridding itself of a requested document. Chambers, 568 F.3d at 1004 (“[A]n agency is.not shielded from liability if it intentionally transfers or destroys a document after it has been requested under FOIA[.]”). The critical issue, then, in a dispute over a document that an agency no longer has, is the agency’s motivation for disposing of or transferring that document. If “the agency is no longer in possession of the document, for a reason that is not itself suspect,” FOIA does not compel the agency “to take further action in order to produce” that document. SafeCard, 926 F.2d at 1201 (emphasis added).
There is no genuine dispute that the Army satisfied that standard. To begin with, the Army’s transfer of documents to the National Archives was done for a proper and eminently sensible reason: to fulfill the Army’s obligations under the Disclosure Act to disclose all relevant materials and “make them available to the public at the National Archives[.]” Pub.L. No. 105-246, § 2(c)(1). That is the antithesis of a suspect motive; following the law is exactly what agencies are supposed to do.
Beyond that, the Army, by complying with the Disclosure Act, already had to declassify and disclose most of the records that DiBacco seeks. Unlike FOIA, the Disclosure Act mandated wholesale disclosure by the agency itself, with no general exemption for classified information and without any request having to be filed or potentially limiting the scope of disclosure. Indeed, the whole point of the Disclosure Act was to spur federal agencies themselves, regardless of any individual request', to declassify and publicly release decades old classified records that had been kept secret on national security and foreign policy grounds. See Interagency Report, at 1; Pub.L. No. 105-246, § 3(a) (defining “ ‘Nazi war criminal records’ ” to “mean[ ] [certain] classified records or portions of records ”) (emphasis added); id. § 3(b)(1) (requiring the Interagency Working Group to “release in their entirety Nazi war criminal records that are described” in the statute, subject to certain exemptions).
*193The Army’s transfer thus bears no col-orable resemblance to FOIA-evasion cases, where an agency tries to thwart disclosure by intentionally moving or destroying responsive documents. See, e.g., Judicial Watch, Inc. v. Department of Commerce, 34 F.Supp.2d 28, 41 (D.D.C.1998) (desig-; nating discovery on whether agency illegally destroyed or discarded responsive information); cf. Chambers, 568 F.3d at 1005 (triable issue of fact on whether agency intentionally destroyed records responsive to a Privacy Act request). Quite the opposite, as a result of the Disclosure Act’s operation, the National Archives has a cache of 1.3 million Army files that it thoroughly searched for records responsive to Oglesby’s FOIA request, netting an additional 2,863 pages of relevant Army records. Murphy Decl. ¶¶ 17-18. The agency has made those records available for public inspection, and DiBacco received copies of them while this appeal was pending. Id.; Letter from Vincent H. Cohen, Jr., Acting United States Attorney for the District of Columbia (April 1, 2015); Letter from James H. Lesar, Counsel for Appellants (April 8, 2015).
DiBacco also attacks the Army’s declarations discussing the transfer and search of documents under the Disclosure Act. She asserts that a discrepancy in the dates attributed to the transfers casts doubt on the declarants’ trustworthiness.
DiBacco is making a mountain out of a molehill. The Army’s declarant, Bradley Dorris, states in one declaration that the Army transferred all World War II files to the National Archives “on or about 26 and 29 September 2000,” First Dorris Decl.. ¶ 4, while in another he indicates a transfer date of “on or about 23 April 2001,” Second Dorris Decl. ¶ 5. The National Archives’ declarant, Martha Murphy, explains that the transfers were completed in phases, with transfers occurring in September 2000, summer 2001, and 2005. Murphy Decl. ¶ 12. The first two dates Murphy provides coincide with the two dates the Dorris declarations reference, which suggests that the Dorris declarations were largely accurate but omitted the last date. That is insufficient to suggest bad faith or dissembling by Dorris or the Army.
More to the point, that discrepancy has nothing to do with the Army’s motivation for the document transfer. What matters is that, under the Disclosure Act, the Army transferred all of its potentially responsive files to the National Archives and did not retain any copies. The Inter-agency Report confirms as much. See Interagency Report, at 54. DiBacco, for her part, is silent as to how publicly available documents at the National Archives— copies of which have always been offered to her for a fee and which she has now in fact received — could plausibly be considered improperly withheld.
Trying another tack, DiBacco argues, for the first time in her reply brief, that the Army violated an Executive Order hy failing to sufficiently declassify information of permanent historical value before transferring the documents to the National Archives. We do not ordinarily consider arguments raised for the first time in a reply brief, and we see no good reason for doing so here. See Abdullah v. Obama, 753 F.3d 193, 199 (D.C.Cir.2014).
At bottom, DiBacco’s main concern is that the transfer allowed the Army to circumvent the fee waiver it granted Oglesby nearly three decades ago. But the Army has assumed, and accordingly so. do we, that Oglesby’s fee waiver extends' to DiBacco, and, following oral argument, the Army provided her with free copies of the 2,863 pages of Army records identified as *194responsive by the National Archives. See supra p. 193. The issue is therefore moot.
That, however, is not the last chapter on those recently released documents. Some were redacted or indicated that pages had been removed, with no accompanying justification for that withholding of information. Letter from James H. Lesar, at 2-3 & nn.1-2. The National Archives’ declaration confirms that some of the documents had redacted information that remains classified. Murphy Decl. ¶ 17a nn.2-3. We accordingly remand to allow the parties to create a record and the district court to decide in the first instance the narrow question of whether those withholdings were permissible under FOIA.
Claims Against the CIA
DiBaceo levels attacks against the CIA’s search for and withholding of responsive records. None has merit.
The CIA’s Search for Records
Like the Army, the CIA maintains that its Disclosure Act search efforts were reasonably calculated to locate documents responsive to Oglesby’s FOIA request. The CIA filed the Interagency Report and four declarations from Martha Lutz, Chief of its Litigation Support Unit, detailing those efforts. First Lutz Declaration (Dec. 14, 2012), J.A. 709-749; Second Lutz Declaration (March 21, 2013), J.A. 1055-1083; Third Lutz Declaration (May 10, 2013), J.A. 1121-1126; Fourth Lutz Declaration (June 20, 2013), J.A. 1147-1156.
The CIA instructed all of its directorates to search for relevant documents, using name and codeword searches. First Lutz Decl. ¶¶ 14-15. With respect to documents concerning Gehlen and his organization, the CIA searched for files retrievable by name, codewords, aliases, and crypto-nyms. Id. The CIA’s searches included operational files exempt from FOIA.
The CIA initially sought and disclosed records “only if they contained either direct information about war crimes or information suggesting that there were grounds to believe that the subject was involved in war crimes, acts of persecution, or looting.” Interagency Report, at 47. That approach led to the declassification and release of approximately 50,000 pages of documents, many of which were redacted. Id. at 49.
In 2005, the CIA adopted the Inter-agency Working Group’s broader interpretation of the Disclosure Act, and accordingly “[d]eclassif[ied] and release[d] information on individuals connected to the Nazis whether war criminals or not,” “[d]eclassif[ied] and release[d] operational project files where Nazis were involved,” and “[u]ndert[ook] additional searches that the [Interagency Working Group] historians or CIA thought necessary.” Interagency Report, at 50. As a result, the agency narrowed redactions in 47,400 pages of the previously released documents, and released over 65,000 new pages. Id. Of relevance here, 2,100 of those pages related to General Gehlen, and the CIA authorized release of another 2,100-page Army file concerning Gehlen. First Lutz Decl. ¶ 13. Continuing its efforts, in May 2012, the CIA provided DiBacco’s counsel with seven discs containing all CIA records released under the Disclosure Act, with information explaining how the records were organized and where records relating to Gehlen and the Gehlen Organization’s relationship with the United States could be located. Id. ¶ 12.
Those efforts discharged the CIA’s FOIA duty to undertake reasonable search efforts. “[A] search need not be perfect, only adequate, and adequacy is measured by the reasonableness of the effort in light of the specific request.” *195Meeropol v. Meese, 790 F.2d 942, 956 (D.C.Cir.1986). The Lutz declarations adequately explain the congruence between the CIA’s Disclosure Act search and Oglesby’s FOIA request. See Second Lutz Decl. ¶¶ 3-4. Lutz identified in detail the locations searched within the CIA, expressly noting that the search included all directorates and even encompassed operational files not subject to FOIA. Id. ¶¶ 6-7.
DiBacco raises a number of specific challenges to the adequacy of the CIA’s search, but none holds water.
First, DiBacco’s arguments regarding the absence of documents concerning secret meetings at Fort Hunt and the failure to use her preferred search terms fail here, just as they did when leveled against the Army. See supra pp. 190-92.
Second, DiBacco points to a December 2000 status report, in which the CIA anticipated locating documents potentially responsive to Oglesby’s request that exceeded the Disclosure Act’s scope. Lutz explained, however, that the statement was based on the CIA’s pre-2005 view that the Disclosure Act did not encompass records pertaining to General Gehlen. Second Lutz Decl. ¶ 4. Five years after filing that status report, the CIA adopted a more expansive view of the Disclosure Act and, under that standard, it reviewed, declassified, and disclosed information on all Nazis (rather than just war criminals), and undertook additional searches that were equivalent to or broader than what FOIA requires. Id. As a consequence, “all Gehlen related records responsive to Oglesby’s request fell within the scope of [the Disclosure Act] and all were released in whole or in part under the [Act] and provided to Plaintiffs.” Id.
Third, DiBacco asserts that there is a discrepancy between the number of responsive documents the CIA predicted in 2000-2001 that it would disclose and what it eventually released. In December 2000 and January 2001, the CIA reported finding “approximately 251 boxes of material, and 2,901 folders” with potentially responsive documents, and that, if the boxes were full, they would likely contain a total of “anywhere between 251,000 and 775,000 pages” of documents. Defendants’ Status Report 1-2, J.A. 622-623; Tr. of Hearing 4:15-23, J.A. 630.
DiBacco seizes on the gap between that prediction and the roughly 115,000 pages the CIA ultimately released. But that differential is no surprise. The CIA based its estimated page count in 2000-2001 on the volume of each box — one cubic foot— and an assumption that all were full of documents responsive to. Oglesby’s request. The boxes had not yet been searched. That the CIA’s later search turned up less than its back-of-the-napkin estimate, which was based on figures and assumptions untethered to the actual contents of the documents, does not impugn the agency’s search.
The CIA’s Exemption Claims
The CIA redacted protected national security information from approximately 475 pages of its disclosure, pursuant to Exemptions 1 and 3, as detailed in its Vaughn index. See First Lutz Decl., Attachment. DiBacco presses global attacks on the use of both exemptions. But neither challenge holds up.
“An agency withholding responsive documents from a FOIA release bears the burden of proving the applicability of claimed exemptions!,]” which it typically does “by affidavit.” American Civil Liberties Union v. Department of Defense, 628 F.3d 612, 619 (D.C.Cir.2011). We “review the district court’s decision on the adequacy of the agency’s showing de novo,” and “must accord substantial weight to an agency’s affidavit concerning *196the details of the classified status of the disputed record.” Id. (internal quotation marks omitted). Summary judgment is warranted based on the agency’s' affidavit if it “describes the justifications for withholding the information with specific detail, demonstrates that the information withheld logically falls within the claimed exemption, and is not contradicted by contrary evidence in the record or by evidence of the agency’s bad faith[.]” Id.
FOIA Exemption 1 excludes from the agency’s general disclosure obligation national defense or foreign policy records “properly classified pursuant to [an] Executive order[.]” 5 U.S.C. § 552(b)(1). In making the Exemption 1 withholdings in this case, the CIA relied on President Clinton’s Executive Order 12958, as amended by President Bush, which was in effect at the time the classifications were made in 2005-2007. See Second Lutz Decl. ¶ 8 & n.6; Exec. Order No. 12958, 60 Fed.Reg. 19,825 (April 17, 1995), amended by Exec. Order No. 13292, 68 Fed.Reg. 15,315 (March 25, 2003). DiBacco maintains that President Obama’s Executive Order currently in effect — Executive Order 13526 — governs. 75 Fed.Reg. 707 (Dec. 29, 2009).
The CIA has it right. A district court may allow an agency to apply a superseding Executive Order during pending FOIA litigation if the agency so requests. Campbell v. Department of Justice, 164 F.3d 20, 29 (D.C.Cir.1998). But “absent [such a request,] the district court may not require the agency to apply the new order; instead, the court must evaluate the agency’s decision under the executive order in force at the time the classification was made.” Id. A court will compel an agency to revisit its classifications only if the superseding Executive Order “calls prior classification decisions” under the earlier order “into question.” Id.
Nothing in the Obama Executive Order calls the CIA’s classification decisions into question. Quite the opposite, the Obama Order explicitly defines properly classified information to include information classified under prior Executive Orders, see Exec. Order 13526, § 6.1(i), and it “does not contain any provision that requires an agency to reconsider classification decisions in pending FOIA litigation,” Campbell, 164 F.3d at 30. To be sure, as DiBac-co observes, the Obama Order significantly changes the automatic declassification process for information that is more than fifty years old. See Exec. Order 13526, § 3.3(h)(1). But what matters for purposes of reviewing the CIA’s classification decisions is that “nothing in the Order requires the district court to apply the new standards in a pending FOIA action.” Campbell, 164 F.3d at 30.
DiBacco separately argues that the withheld documents lack markings required to properly classify documents under the Clinton/Bush Order. That too is wrong. The documents at issue were classified decades ago, with the markings required under the then-governing Executive Orders. See Fourth Lutz Deck ¶¶ 8-9 & n.9. That is all the Clinton/Bush Order requires. Exec. Order 12958, amended by Exec. Order 13292, § 1.6(f) (“Information assigned a level of classification under this or predecessor orders shall be considered as classified at that level of classification despite the omission of other required markings.”).
DiBacco also challenges the CIA’s reliance on Exemption 3, which generally excludes from disclosure matters that are “specifically exempted from disclosure by statute.” 5 U.S.C. § 552(b)(3). The CIA relied on the National Security Act of 1947, 50 U.S.C. §§ 3001 et seq., to justify withholding information that would reveal *197intelligence sources and methods. See First Lutz Decl. ¶¶ 30-32, 35-51.
“Exemption 3 differs from other FOIA exemptions in that its applicability depends less on the detailed factual contents of specific documents; the sole issue for decision is the existence of a relevant statute and the inclusion of withheld material within the statute’s coverage.” Morley v. CIA, 508 F.3d 1108, 1126 (D.C.Cir.2007) (internal quotation marks omitted). The National Security Act provision invoked by the CIA provides that “[t]he Director of National Intelligence shall protect intelligence sources and methods from unauthorized disclosure.” 50 U.S.C. § 3024(i)(1).
DiBacco does not dispute, nor could she, that Section 3024(i)(l) is a valid Exemption 3 statute. Sims, 471 U.S. at 167, 105 S.Ct. 1881. Nor does DiBacco challenge the CIA’s determination that the withheld material contains “intelligence sources and methods” within the National Security Act’s coverage.
Instead, DiBacco focuses on language in that same statutory subsection providing that “[t]he Director [of National Intelligence] may only delegate a duty or authority given the Director under this subsection to the Principal Deputy Director of National Intelligence.” 50 U.S.C. § 3024(i)(3). That language was added by the Intelligence Reform and Terrorism Prevention Act of 2004, Pub.L. No. 108-458, § 102(A)(i)(1), 118 Stat. 3638. Prior to 2004, the Director of Central Intelligence bore responsibility for protecting intelligence sources and methods, and had the authority to invoke the National Security Act to prevent the unauthorized disclosure of such information under FOIA. See Sims, 471 U.S. at 167, 105 S.Ct. 1881.
DiBacco argues that the 2004 amendment stripped the CIA of that power, at least absent an express authorization from the Director of National Intelligence. Although DiBacco’s reading of the amended statute is not entirely clear, she does concede that the Director of National Intelligence may delegate the authority to protect intelligence sources and methods to the Director of the CIA. Oral Arg. Rec. 13:20-13:30. She asserts, however, that the delegation must be done on a case-by-case basis and that the absence of such a particularized delegation here dooms the CIA’s Exemption 3 claims. Id. at 13:30-15:50.
DiBacco’s argument misunderstands the governing statutory scheme. First, even if the Director of the CIA needs authorization to protect intelligence sources and methods from unauthorized disclosure, that authorization is already in place. Both the President and the Director of National Intelligence have provided it. In Executive Order 12333, as amended, President Obama ordered the Director of the CIA to “[p]rotect intelligence and intelligence sources, methods and activities from unauthorized disclosure in accordance with guidance from the Director [of National Intelligence].” Exec. Order 12333, 46 Fed.Reg. 59,941 (Dec. 4, 1981), amended by Exec. Order 13470, § 1.6(d), 73 Fed.Reg. 45,325, 45,332 (July 30, 2008). Pursuant to that Executive Order and the National Security Act, the Director of National Intelligence has issued such guidance, ordering heads of intelligence agencies — such as the Director of the CIA — to “[p]rotect national intelligence and intelligence sources, methods and activities from unauthorized disclosure[.]” Intelligence Community Directive 700, at 3 (June 7, 2012), available at http:// www.dni.gov/files/documents/ICD/ICD_ 700.pdf.
Second, statutory language must always be read “in [its] context,” and that context *198supports the ability of the Director of National Intelligence to delegate to the CIA Director subject to the former’s guidance. King v. Burwell, — U.S. —, 135 S.Ct. 2480, 2492-93, 192 L.Ed.2d 483 (2015) (internal quotation marks omitted); see Association of Civilian Technicians, Inc. v. United States, 603 F.3d 989, 992 (D.C.Cir.2010) (“[WJords are to be read in the context in which they are used and in the broader context of the statutory scheme.”). The paragraph that precedes the delegation limitation mandates that “the Director of National Intelligence shall establish and implement guidelines for the intelligence community” addressing, among other things, the “[cjlassification of information” and “[ajecess to and dissemination of intelligence,” “[cjonsistent with” the Director’s duty to protect intelligence sources and methods. 50 U.S.C. § 3024(i)(2)(A)-(B). The statute thus expressly envisions the National Intelligence Director giving guidance to intelligence agencies, which necessarily is guidance for those agencies to use.
The Director of National Intelligence has exercised that authority. Intelligence Community Directive 700 establishes the intelligence community’s “policy for the protection of national intelligence” and provides a “framework” for oversight of classified information and “protection of national intelligence and intelligence sources, methods, and activities.” Intelligence Community Directive 700, at 1. By ordering the heads of components of the intelligence community to “[pjrotect national intelligence and intelligence sources, methods and activities from unauthorized disclosure,” id. at 3, the Director of National Intelligence exercised his power to issue guidelines in a manner “consistent with” his statutory duty, 50 U.S.C. § 3024(i)(2).
Third, the National Security Act’s structuring of the intelligence community likewise confirms the authority of the Director of National Intelligence to provide rules and guidance for intelligence agencies to implement on a case-by-case basis. The Act provides that, “[sjubject to the authority, direction, and control of the President,” the Director of National Intelligence “serve[sj as head of the intelligence community” and “overseefsj and direct[s] the implementation of the National Intelligence Program,” 50 U.S.C. § 3023(b), which consists of “all programs, projects, and activities of the intelligence community,” id. § 3003(6). The statute thus expressly contemplates that the Director of National Intelligence, under the President’s direction, will issue general directives that control the manner in which the intelligence community as a whole carries out its mission.
Underscoring the point, the provision of the National Security Act listing the Director of National Intelligence’s “[rjespon-sibilities and authorities” is chock full of provisions tasking the Director with formulating and issuing guidance to govern the intelligence community writ large. 50 U.S.C. § 3024. Notably absent are provisions suggesting that the Director must— or even feasibly could — have his fingers perpetually in the day-in-and-day-out operations of each and every one of the sixteen components of the intelligence community.4 To the contrary, the National Security Act assigns the CIA Director responsibility for the day-to-day conduct of the agency’s mission. See 50 U.S.C. § 3036(b)-(f).
Finally, we would require far more explicit statutory direction before concluding that Congress meant to saddle the highest-*199level official in the intelligence community (other than the President) with such micromanagement, or meant to so ossify the ability of the intelligence community to protect its most vital intelligence information. DiBacco’s interpretation “overlooks the practical necessities of modern intelligence gathering,” and would improperly naiTOW the Director of National Intelligence’s “very broad authority to protect all sources of intelligence information from disclosure.” Sims, 471 U.S. at 168-169, 105 S.Ct. 1881. Nothing in the 2004 amendment to the National Security Act indicates a congressional intent to hamstring the Director in that fashion.5
Further, the overall scheme for protecting such sensitive information leaves it to the President to dictate the duties (in addition to those statutorily enumerated) of the Director of National Intelligence, and the President and the Director of National Intelligence do the same for the Director of the CIA. See 50 U.S.C. § 3024(f)(8) (“The Director of National Intelligence shall perform such other functions as the President may direct.”); id. § 3036(d)(4) (The Director of the CIA’s statutory responsibilities include “such other functions and duties related to intelligence affecting the national security as the President or the Director of National Intelligence may direct.”).
Executive Order 12333, in turn, requires the Director of the CIA to protect intelligence sources and methods from unauthorized disclosure in accordance with guidance issued by the Director of National Intelligence. Exec. Order 12333, amended by Exec. Order 13470, § 1.6(d). The Order also requires the Director of National Intelligence to “ensure that programs are developed to protect, intelligence sources, methods, and activities from unauthorized disclosure.” Id. § 1.3(b)(8). Intelligence Community Directive 700 provides the guidance contemplated by Executive Order 12333, and likewise gives the Director of the CIA the duty to protect intelligence sources and methods from unauthorized disclosure. The CIA Director’s exercise of that authority consistent with the National Intelligence Director’s guidance falls naturally within the “other functions and duties related to intelligence” that the President or the Director of National Intelligence may grant the Director of the CIA under the National Security Act. 50 U.S.C. § 3036(d)(4).
Accordingly, when read in context, the statutory limitation on delegation on which DiBacco relies, 50 U.S.C. § 3024(i)(3), does not unravel either the President’s or the Director of National Intelligence’s authority to assign responsibility to intelligence agency heads to protect intelligence sources and methods. Instead, the anti-delegation provision means that the Director of National Intelligence must hold close those critical responsibilities for superintending and guiding the work of members of the intelligence community. The Director has done that through the guidance issued, and DiBacco does not dispute that the CIA made its withholding decisions in this case under the framework that guidance provides.
DiBacco’s Motion to Compel
DiBacco makes a glancing challenge to the district court’s denial of her motion to compel disclosure of the Army’s and CIA’s *200classified declarations referenced in their summary-judgment papers filed nearly two decades ago. DiBacco’s argument focuses on when a district court may rely upon such declarations to decide FOIA exemption claims. See Plaintiffs’ Motion for an Order Compelling Defendants Central Intelligence Agency and the Department of the Army to Disclose Ex Parte Declarations 4-6 (April 25, 2013), ECF No. 249.
But the district court never ruled on the agencies’ earlier motion for summary judgment because they withdrew it after the 1998 passage of the Disclosure Act. Those declarations thus played no role in resolving that summary judgment motion, nor did the district court rely on them in deciding any other issue. On top of that, the district court concluded that the unclassified declarations filed in the case were sufficient to enable DiBacco to oppose, and the court to resolve, the agencies’ current motion for summary judgment. DiBacco, 983 F.Supp.2d at 53. DiBacco has not challenged that conclusion. Nor has she provided any reason to disturb the district court’s exercise of its “broad discretion regarding whether to conduct in camera review,” let alone to second-guess its refusal to disclose classified declarations that it appropriately declined to review. Larson, 565 F.3d at 870.
Ill
Conclusion
We affirm the district court’s grant of summary judgment to the Army and CIA with respect to (i) the Army’s transfer of documents to the National Archives, (ii) both agencies’ searches for responsive documents, and (iii) the CIA’s withholding of information under Exemptions 1 and 3. Our remand is limited to issues arising 'from the Army’s release to DiBacco during the appeal of responsive but redacted Army documents that had been held by the National Archives.
So ordered.