We’re learning more about how that enormous cache of telephone metadata at the National Security Agency is actually used. According to two sources in the intelligence community who have worked with the system, it’s one of many tools available to analysts working on terrorism investigations or providing intelligence for military forces overseas.
One former defense intelligence employee describes it this way:
The NSA makes a list of names and/or phone numbers available to analysts who are cleared to use the meta database. These names and/or numbers have been obtained by NSA through other collection programs, presumably legal ones. The analysts input those names and/or numbers to the meta database, which will then show any connections to phone numbers in it.
The meta database itself doesn’t contain any names—it is only phone numbers. If there’s a number that’s based in the United States, the analyst only sees an “X” mark. If he wants to see the number underneath that X, he has to get clearance from higher authority—in this source’s experience that was the general counsel of the organization where he worked.
The source also said the tool wasn’t particularly useful. It’s there to help analysts better understand the links between potential terrorists, and to help identify them. But the analyst said that searching for numbers and names on Google often led to better results.
An intelligence official who has used the meta dabtase confirmed the description of how it works. But he said the presence of so many “innocent” numbers in the system posed a challenge. Analysts have to weed through them to find only the numbers they’re allowed to see without permission from a higher authority.
The meta database is one of dozens of different systems or intelligence streams available across the intelligence agencies. From the sources’ descriptions, it sounds relatively mundane compared to the other tools that are available.
However, one of those other tools, which was revealed yesterday by the Guardian and the Washington Post, called PRISM, appears far more secretive and less widely used. Neither of these source had ever heard of it. The former defense intelligence employee expressed alarm that, according to reports, the system gives the NSA direct access to the central servers of some of the country’s biggest Internet companies, including Facebook and Yahoo!, and then lets analysts obtain e-mails, video and audio files, photographs, and documents.
This individual said he couldn’t explain how, based on his training and experience, the PRISM system complies with the law. It doesn’t seem to be discriminate enough in separating US person’s content—such as their e-mails—from those of foreigners. The government almost always needs a warrant to look at content.
Reportedly, PRISM sweeps up the information of US persons when analysts tap into those central servers. They’re instructed to document these “incidental collections,” but are told, according to a training manual reviewed by the Post, that “it’s nothing to worry about” if Americans are caught up in the stream.
The intelligence official said that based on reports, the PRISM system would have to be collecting massive amounts of information, and that NSA was likely the only agency with the computing power and the storage space to handle it all. The agency has been running out of electronic storage at its Ft. Meade, Md., headquarters and has built a new 1-million square foot data center in the Utah desert.