clock menu more-arrow no yes

Filed under:

NSA reportedly collects only 20 to 30 percent of call records, doesn't cover 'most' cellphones

New, 32 comments

As the FISA court approves limits to the NSA's phone-record tracking, anonymous sources tell The Wall Street Journal that the agency actually collects 20 percent or less of American call records, contradicting reports that the program creates a comprehensive database. According to people apparently familiar with the matter, the agency has struggled to adapt to the decline in landline use, and the orders sent to companies don't cover "most" mobile phone records. One official says that because the agency is not allowed to keep cellphone location records in the database, figuring out how to purge them has delayed attempts to modernize the program. NSA head Keith Alexander has said that the agency is not currently collecting any location information under the program, but in 2010 and 2011, it gathered cellphone location data as part of a test project.

The Washington Post has published similar reports, although officials tell it that the number is between 20 and 30 percent. According to them, however, this is the result of a steep decline: in 2006, the NSA was reportedly collecting "nearly all records from a number of US companies," but the number had dropped to 30 percent by the summer of 2013, when the program's existence was first leaked. Officials tell the Post that the agency has been attempting to bring that number back up and is preparing to ask more phone providers for records; in some cases, agents can also effectively get information that's not in the database when people on non-covered carriers call a network whose records are in the database.

A limited and largely landline-based database would greatly reduce the chances of surveillance for many Americans, though the total number of records would still be vast. As the Journal notes, a line in one of the reviews commissioned by the White House states that "the metadata captured by the program covers only a portion of the records of only a few telephone service providers." Former CIA deputy director Michael Morell, the panel's head, also wrote an editorial in which he said the database covers "only a subset of the total calls made." Given how few service providers operate in major US markets, the former claim is somewhat ambiguous. A surveillance program that collects less than a quarter of records and focuses on landlines, however, seems to clash with the intelligence community's own description of what the database allows.

In his editorial, Morell said that the tool could provide evidence that "certain known terrorists were most likely not in phone contact with anyone in the United States." Director of National Intelligence James Clapper has made similar claims, and the deputy attorney general, James Cole, said in July of 2013 that "if you're looking for the needle in the haystack, you have to have the entire haystack to look through." Another White House report said that one of the benefits of the program was negative reporting. "Analysis of telephone calling records can establish that a known terrorism suspect overseas has not been in telephone contact with anyone in the United States," it says — an assurance that becomes much less useful as the number of records drop.