Is There a Sustainable Ops Tempo in S2? How Can Analysts Deal with the Flood of Collection?

Open PDF
Creating agency: NSA National Security Agency (US)
Document date: 2011-04-06
Type: SIDtoday
Pages: 3
Genre: newsletter
Classification: SECRET SI
Dissemination: REL

The NSA's Deputy Director for Analysis and Production discusses the unsustainable operational tempo caused by overwhelming data collection volumes, advocating for "Analytic Modernization" that would shift from treating SIGINT as "radioactive" NSA-only material to sharing raw intelligence with other agencies for collaborative exploitation.

Long Description:
This April 2011 SIDtoday article presents the conclusion of an interview with the SID Deputy Director for Analysis and Production, focusing on why the NSA has failed to create a sustainable operational tempo despite previous efforts. The official acknowledges that analysts can only process a "fractional portion" of available intelligence and argues for fundamental changes in how NSA handles SIGINT data. The proposed "Analytic Modernization" represents a philosophical shift away from decades of treating signals intelligence as "radioactive" material requiring careful NSA management under Fourth Amendment protections. The interview reveals staggering data volumes overwhelming NSA systems: 2 petabytes stored at any given time (equivalent to 20 million four-drawer filing cabinets) with projections of 1 terabyte per second by year's end. The official advocates offloading non-cryptologic analysis to Intelligence Community partners and customers, using examples like immediately sharing terabytes of captured CAD weapons designs with government weapons experts rather than having NSA analysts process them. This approach would involve storing data in shared repositories like the Library of National Intelligence and using collaborative platforms like A-Space, shifting compliance responsibilities outside NSA while allowing analysts to focus on uniquely cryptologic problems. The official also notes emerging challenges including the decline of traditional "strong selectors" like persistent phone numbers, increased commercial encryption, and the need for pattern-of-life analysis rather than traditional targeting methods.

(S//SI//REL) Is There a Sustainable Ops Tempo in S2? How Can Analysts Deal with the Flood of Collection? -- An Interview with [REDACTED] (conclusion)

FROM: the SIDtoday Editor
Run Date: 04/06/2011

(U//FOUO) Editor's note: Here's the conclusion of our interview with SID's Deputy Director for
A&P [REDACTED] (pictured). (If you missed part one, you can find it HERE.) In this segment
[REDACTED] comments on Analytic Modernization and why it is taking center stage...

  1. Q: (U) When we last interviewed you in August 2009 -- you were Assistant DDAP
    at the time -- you spoke of the importance of creating a sustainable ops tempo in S2.
    Have you succeeded?

A: (U). No. No, I haven't... And it's not just because there's so much happening in the world. Here's
the thing: we have to take a fundamentally different approach to how we do business. The volume
and quality of access we have now gives us an unprecedented capability to produce some of the
best intelligence we ever have, but the fact is that we are only getting to a fractional portion of what
we have access to. I can't tell you for sure that as good as what we produce is, that it's the absolute
best we can do. It's not sustainable for the workforce or the mission to keep working in the way we
always have.

(U//FOUO) We've embarked on Analytic Modernization. In the past there have been
transformations... modernizations... renovations... it gets tiring! But to a large degree those efforts
have represented supercharging existing capabilities -- doing more with less. We need to
fundamentally change how we interact with SIGINT.

(U//FOUO) This is a big deal, because for decades we've trained our workforce -- and took pride in
the fact -- that we had the unique responsibility within the IC to manage this information in
accordance with the 4th Amendment. To that extent we thought of SIGINT as "radioactive" and it
was our job to render the information safe for others. Our process was about managing the
information in a manner consistent with the privacy rights of U.S. persons.

(U//FOUO) I believe we have developed an awareness of the information and an understanding of
our authorities that allows us to think differently about that relationship... and in so doing, to create
an increased capacity and recover more time for the analysts. I don't mean only analysts in SID, but
also analysts in IAD, in NTOC, etc. The key is to better leverage the Intelligence Community and
our partner relationships for the exploitation of SIGINT, which is something in the past we would
not have done to the degree we're proposing -- it was considered our domain (and "radioactive").

(U//FOUO) We must take advantage of the expertise and capabilities in the IC and our customer
base to enhance discovery and capacity, and to make the actionable information available almost as
soon as we encounter it. This is not about turning NSA into a collection resource for others
(although collection is in fact one of our great strengths and one we are uniquely qualified to
undertake) -- it's about making sure we don't expend cryptologic resources doing work that is not
uniquely cryptologic in nature -- work that others can do.

(U) Q: Can you give an example?

A: (S//SI//REL) If NSA discovers a pathway into an adversary's information space and we extract a
terabyte of CAD [computer-aided design] drawings of weapon designs, is it of best value for the IC
to have talented NSA analysts work their way through this 1 terabyte of data? Is it uniquely
cryptologic? Or is it of better benefit to the nation that we expose it immediately to the best and
brightest weapon designers in the US government to work on in a collaborative space, to triage,
assess and exploit the value of that information?

(U//FOUO) In so doing we are not precluding NSA analysts from continued access to the data, any
more than a published SIGINT product report isn't available for future reference. We are just
leveraging the power and expertise of others so that we can turn our attention to that that which
only we can exploit by virtue of our unique talents.

Q: (U) Isn't it possible that our customers won't want to devote their own people
and time to analyzing the data?

A: (U//FOUO) If we get access to the best information on a topic, the value will speak for itself and
the customers will make the resources available. It's like YouTube videos -- they "go viral" when
people with good reputations recommend a video to others. If we put the information out there and
monitor the customer response to it, we'll know when to recommend specific items they might
want to take a close look at. Amazon does the same thing by looking at how you react to the
products you've looked at. We can figure out what the communities of interest are for that topic... If
the material is in a foreign language, we have tools they can use to get the gist of it, and if it looks
promising, they can use the National Virtual Translation Center to get it translated. We want our
language analysts focused on uniquely cryptologic problems that rarely boil down to a
straightforward translation.

(U//FOUO) So what does this approach accomplish? We've exposed the intelligence to people who
can interpret it and use it, and we've created opportunities for collaboration. We've also off-loaded
the responsibility to manage the data in our respositories and own the compliance responsibility. If
data is stored in the Library of National Intelligence, someone else is paying for it.

(U//FOUO) In the end, we exist to produce information. The only way to go more deeply into
targets is to avoid getting stuck on production that others can do for themselves. Our challenge is to
be always out looking for something new. We need to think about problems, not just about
production.

(U//FOUO) The collaborative component means it's not an NSA view, a DIA view, etc... Rather, it's
opened up to all on A-Space. I would like to see people log onto A-Space and announce "I want to
share traffic and create a multi-seal report based on joint input." Why not collaborate at the point of
discovery?

Q: 7. (U//FOUO) Various pushes for analytic modernization have been going on
for decades at NSA, but now the issue really seems to be taking center stage. In
fact, the number one "SIGINT Goal for 2011-2015" is to "revolutionize analysis."
What's different now?

A: (S//SI//REL) We live in an Information Age when we have massive reserves of information and
don't have the capability to exploit it. I was told that there are 2 petabytes of data in the SIGINT
System at any given time. How much is that? That's equal to 20 million 4-drawer filing cabinets.
How many cabinets per analyst is that?? By the end of this year, we'll have 1 terabyte of data per
second coming in. You can't crank that through the existing processes and be effective.

Q: (U) ...So it's a matter of volume?

A: (S//SI//REL) Not volume alone, but also complexity. We need to piece together the data. It's
impossible to do that using traditional methods. Strong selectors -- like phone numbers -- will
become a thing of the past. It used to be that if you had a target's number, you could follow it for
most of your career. Not anymore. My daughter doesn't even make phone calls, and many targets
do the same. Also, the commercial market demands privacy, and this will drive our targets to go
encrypted, maybe into unexploitable realms. Our nation needs us to look for patterns surrounding a
particular spot on Earth and make the connections -- who can do that if not us? And we can't do it
using traditional methods.

  1. Q: (U) Looking into the future, is there anything that especially worries
    you? ...An eventuality (internal or external) that would make it hard for A&P to
    continue to put out quality intelligence?

A: (U//FOUO) I'm worried that we have so much good stuff that we could lock down analysts and
have them just producing product, and something would jump out and surprise us. So we need the
discipline to invest in the wild and the unknowns.

  1. Q: (U) Is there anything else you'd like to say to the SIGINT workforce?

A: (U) At the risk of being melodramatic, as hard as the work is and as challenging as the work
environment can be, we should never forget how lucky we are to have a job where we can have so
much impact for the nation. Don't let the turkeys get you down!