Update:
Sorry, I didn't see this was diaried intensely by The Free Agent. I had no time to read through the material here at dailykos this morning and just wrote this spontaneously the moment I became aware of it. I am glad others covered it so well.
So I run through my morning duties and stumble into this article from the Atlantic Wire:
The NSA Admits It Analyzes More People's Data Than Previously Revealed.
That headline caught my interest and I read among others this:
Chris Inglis, the agency's deputy director, was one of several government representatives—including from the FBI and the office of the Director of National Intelligence—testifying before the House Judiciary Committee this morning ...
But Inglis' statement was new. Analysts look "two or three hops" from terror suspects when evaluating terror activity, Inglis revealed. Previously, the limit of how surveillance was extended had been described as two hops. This meant that if the NSA were following a phone metadata or web trail from a terror suspect, it could also look at the calls from the people that suspect has spoken with—one hop. And then, the calls that second person had also spoken with—two hops. Terror suspect to person two to person three. Two hops. And now: A third hop.
Of course those "hops" made me jump and I searched for the transcript of that Judiciary Committee Hearing yesterday.
As google would say "I got lucky". We had not only the hearing's transcripts but also the complete video footage. I can't share the video footage I am looking at and have no time to search if it's available online somewhere. First, let me give you how the Atlantic Wire explained it:
Let's say the government suspects you are a terrorist and it has access to your Facebook account. If you're an American citizen, it can't do that currently (with certain exceptions)—but for the sake of argument. So all of your friends, that's one hop. Your friends' friends, whether you know them or not—two hops. Your friends' friends' friends, whoever they happen to be, are that third hop.
As is explained in this article of the NYT:
Separating You and Me? 4.74 Degrees, we are only 4.74 degrees separated from each other on the intertubz (my, why are people still lonely these days, I ask).
I mean not even six degrees of separation, only 4.74 and what does the little difference between 4.74 to t degrees really mean? A couple of billions people less surveileed, that's just peanuts.
I searched the committee's transcript for the part where Chris Inglis said what is described in the Atlantic Wire article. It was an exchange between Barbara Jackson Lee and Inglis. Here it goes.
JACKSON LEE: ....
So I pose the first question that deals with the idea that witnesses have testified in recent hearings that the phone-record data were queried at 300 times last year. How do you define a query, and how do you define the necessity of what I call "trolling?
And someone wanted to have me rephrase that. But the gathering of millions and millions of meta-data gathering, how do you define "query" first, but then how do you justify that gathering?
INGLIS:
Yes, ma'am. I'll take that question. So first, the court has approved procedures by which we can form a selector. The reasonable articulable suspicion standard was what we described earlier. And less than 300 times in 2012 we approved a selector for entering the database.
The court also approves what's called...
JACKSON LEE:
The query is based upon permission by the FISA court?
INGLIS:
Yes, ma'am. The FISA court approves the rules. But as we've described in this hearing, the decisions about how to form those selectors are made at the National Security Agency, and subject to auditing and review.
JACKSON LEE:
So the query is made without a warrant. You go by criteria that's been set, and then you make a query, and a preliminary oversight, if you will. Is that what you're saying?
INGLIS:
That's correct, ma'am. And can I just then add that the court has also given permission to do, not just first-hop analysis, meaning what numbers are in contact with that selector, but to then from those numbers, go out two or three hops.
In many of the cases that Ms. Douglas referenced earlier, it was at the second hop. It was at that second connection that something of interest came that then caused the Federal Bureau of Investigation to apply their resources to essentially uncover or add additional information to terrorist activity.
JACKSON LEE:
Once you do the query out of the 300, then what are the next step?
INGLIS:
So that query, when it's returned, can be a first-hop query, or a second- or a third-hop query. That information is then reviewed by the National Security Agency analyst. And a report would be written and disseminated to the Federal Bureau of Investigation if we see something that would be of interest to them.
In many cases, when a query is performed, nothing of consequence turns up. No connections that are (inaudible) turn up, and if were no report would be made.
I hope they don't find a text message from someone who just had bought some skittles ...
Ok, guys, as mentioned in one of the twitter messages within the Atlantic Wire article, the best thing is to know nobody and never talk to anybody. We can actually just bury ourselves alive to make sure, we didn't inadvertently endangered a person to become a suspect just because ... we are already dead anyhow.
PS. All bolds are mine and I won't be here to comment. But like to read your's. Have to jump from hops to hops to satisfy my employer's curiosity.