The “Automated Query” at the Telecoms Will Include “Correlations”

In addition to Mike Rogers’ confirmation that HPSCI does not intend HR 3361 to change any of the voluminous collection programs the intelligence community does aside from the phone dragnet, his report on the bill also drew my attention to this previously public detail I had overlooked.

3 The Committee understands that ‘‘[t]he first ‘hop’ from a seed returns results including all identifiers (and their associated metadata) with a contact and/or connection with the seed. The second ‘‘hop’’ returns results that include all identifiers (and their associated metadata) with a contact and/or connection with an identifier revealed by the first ‘hop.’’ ’ In re Application of the FBI for an Order Requiring the Production of Tangible Things, BR 14–01, at 1–2 n.1 (FISC Feb. 5, 2014). [my emphasis]

This is a description of the currently desired “hop” system (though not, I don’t think, what is fully in place) connecting people through their phone — and likely, other communications — habits.

Before I get into what it says, let’s look at where it points. The language here is from a footnote on page 14 of the bill report–suggesting it’s something Mike Rogers wanted to make sure got in the Legislative Record. It cites back to the February 5, 2014 order amending the January 3 order to include the Administration’s request to have FISC review all the query terms.

I don’t believe (but could be wrong — the new FISC docket is far less usable) that we ever got the revised order. But in the order to amend the order also dated February 5, that language appears in footnote 3. The footnote itself cites to the original application for the order dated January 3. But the reference footnoted cites the January 3 order, page 11-12. The footnoted discussion is a part (or summary) of the entirely redacted description of the automated query starting on page 11 and taking up all of page 12 of the order.

That is, this language on hops provides an unclassified version of the classified description of the automated query process (the one they haven’t gotten running yet).

So this is (part of) what the government has been trying — but failing, since November 2012 — to get up and running.

Which is reportedly one of the reasons the Intelligence Community has decided it may be in their best interest to outsource this to the telecoms.

In other words this language provides clues about why the IC was willing to outsource the dragnet.

The description of the hops reveals two things that got added to the 3- or 2-hop process the government once described.

First, they’re including “associated metadata” among the things that can be further chained. Even assuming we’re only talking voice telecom information, this would include cell site location on top of the other metadata (and, in the era of smart phones, potentially far, far more).

But in addition, they’re including “connections,” in addition to contacts, with the seed.

That is, you don’t have to ever call a target to be sucked up in the phone dragnet. You can be simply “connected” to that target. The kinds of connections in question surely include dropped burner phones (that is, a matching of phones that call the same pattern of phones as an inactive phone, and therefore are really targeting the same person). They may include common geolocation. But — again, given the advent of smart phones — they could include far, far more.

So what this little footnote calls to my attention (thanks, Mike Rogers!) is that they’ve gotten approval for different kinds of chaining, beyond actual phone contacts (remember, this could include Internet contacts over a smart phone). And they’ve included metadata generally, not just phone call records, surely including geolocation, among the things they might chain on.

Which explains one incentive for outsourcing this. They can’t use geolocation for chaining in government hands. They can in private hands. There’s likely far more information for which that is true when you consider smart phones.

They can’t access that information now. They will be able to once HR 3361 outsources everything to the telecoms.

But really, this is about reform.

Update: This post was tweaked on 5/18 for clarity.

Tweet about this on Twitter0Share on Reddit0Share on Facebook0Google+0Email to someone

14 Responses to The “Automated Query” at the Telecoms Will Include “Correlations”

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10
  • 11
  • 12
  • 13
  • 14
Emptywheel Twitterverse
emptywheel @onekade No no. Other stuff. Very pressing. Tomorrow. Also, stupid NGO fed me some actually interesting 12333 documents so mouth watering.
15mreplyretweetfavorite
emptywheel @onekade Oh yeah.
16mreplyretweetfavorite
emptywheel @joanneleon You've been sitting at home with a bowl, testing? Consumer Reports might pay for that. @NamelessCynic
17mreplyretweetfavorite
emptywheel @joanneleon Right. Basic math right? More surface area more crappy cheap chocolate. @NamelessCynic
17mreplyretweetfavorite
bmaz @emptywheel @ShitSquare Whaaaa? Who, me?
19mreplyretweetfavorite
emptywheel RT @GRPress: Surfers braved big waves and snow today in Muskegon. That's Pure Michigan. Video: http://t.co/qFhFHfmWo8 http://t.co/YjCCK8DKaf
20mreplyretweetfavorite
emptywheel @bmaz Hey, get off my @ShitSquare and start trash talking me about Noodle Arm storming Foxboro.
20mreplyretweetfavorite
bmaz Bingo. RT @emptywheel This is why Obama prolly could get Preet confirmed AG. http://t.co/LUV3herhhv
21mreplyretweetfavorite
emptywheel @joanneleon Too much oxygen, too little nested treasure of HCFS nougatty goodness. Also, fewer peanuts. @NamelessCynic
21mreplyretweetfavorite
bmaz RT @FeleciaForAZ: "It's clear to me that APS is not making an endorsement, it's paying off Mr. Brnovich and he is welcoming this... http://…
22mreplyretweetfavorite
emptywheel @nickmanes1 Sadly for Chuck Schumer he left open the "best Wall Street owned Senator on Twitter" role. @TimFernholz
25mreplyretweetfavorite
emptywheel @nickmanes1 Booker would be Wall Street's baby even if he didn't live in NJ. @TimFernholz
28mreplyretweetfavorite
May 2014
S M T W T F S
« Apr   Jun »
 123
45678910
11121314151617
18192021222324
25262728293031