Page 83 - pclob usa freedom

TOP SECRET//SI//NOFORN

resulted in the issuance of only 15 intelligence reports. While we may not expect metadata
collection activities to produce as many reports as content collection activities, the government
itself noted the program’s limited relative value. 357 The Board was informed, moreover, that FBI
found the reports largely (though not wholly) redundant: in only two instances did FBI receive
unique information from USA Freedom Act CDRs.

(TS//SI//NF) Some of the reasons the CDR program did not produce a large volume of
useful intelligence can be traced back to evolutions in technology since the first iteration of a
post-9/11 telephony metadata program. Experts in and out of government have noted a shift
away from traditional telephony, with terrorists increasingly using chat applications and
encrypted messaging. Yet CDRs collected under the USA Freedom Act did
. As a result, the program could not have
detected the 2019 analog of the reason it was created: to see
The blame does not rest with NSA.
Traditional telephone records simply do not carry the same importance they once did; no version
of a domestic metadata program fixed solely on traditional telephony was likely to have
produced intelligence reflective of its costs.

(U) That should not, however, distract us from the reality that multi-hop analysis can
have important intelligence value. Simple commonsense examples illustrate how. First,
consider the case of a terrorist organization using a trusted intermediary, or “cutout,” for
communications. The government may be investigating a particular target (“A”) who
communicates with a person (“B”), who in turn communicates with a senior terrorist leader
(“C”). Investigators know of both A’s and C’s connection to the terrorist network, but do not
know with whom C is communicating. If investigators are unaware of B, single-hop collection
would show only that A is communicating with an unknown party, B. Because B is unknown,
the government may be unable to establish the “reasonable articulable suspicion” necessary to
secure process for further hops. 358 Two-hop collection, by contrast, allows the government to
see that A and C may be communicating through an intermediary, thereby bringing B’s potential
significance as a cutout to investigators’ attention.

(U) This is but one of many examples that illustrate how contact-chain analysis may add
significant intelligence value to national-security investigations. Here are two more. Consider
that the target of an investigation (“A”) could be communicating directly with a senior terrorist
leader (“C”) and also, at the same time, with an unknown party (“B”). The government may

(U) 50 U.S.C. § 1861(c)(2)(F)(iii)–(iv). For ease of exposition, we use the more colloquial term “hop,” but we mean
to capture the precise text of the statute.
357 (U) See NSA briefing to the Board (May 23, 2019); see also Letter from Daniel Coats, Director of National
Intelligence, to Senators Richard Burr, Lindsey Graham, Mark Warner, and Dianne Feinstein (Aug. 14, 2019).
358 (U) 50 U.S.C. § 1861(b)(2)(B).
80

TOP SECRET//SI//NOFORN