Alex July 11, 2013 6:55 AM

The Washington Post disclosed Saturday that the existence of a top-secret NSA program called NUCLEON, which “intercepts telephone calls and routes the spoken words” to a database.

SJ July 11, 2013 7:53 AM

In mostly-unrelated news, a private-sector corporation that I’ve worked for has specifically asked employees not to disclose internal project names on Linked-In profiles.

Employees were encouraged to describe skill-sets and generic tasks, but discouraged from describing specific projects or tasks.

This may be something that the corporate world is still figuring out. Is the management structure of government agencies also learning about this?

I’d be very surprised to learn that the NSA doesn’t already have a rule set forbidding release of those codenames into outside-of-NSA publications.

Michael July 11, 2013 8:22 AM

The “rule set forbidding release” of ‘codenames’ is mostly handled through classification guidance.

For a most part, a name, acronym, or initialism will be unclassified to facilitate communication but it becomes classified when associated with a specific function or capability. For example, a conversation could take place over an open phoneline and reference ‘SQUIRRELFIRE’ as long as neither party doesn’t mention what that is or does.

And names occasionally change due to leaks.

S July 11, 2013 8:24 AM

And the linked to linkedin profile is still there! somebody will get his ass spanked…. :-).

Frank Wilhoit July 11, 2013 8:26 AM

If there is any poetic justice left in the world (as there certainly is none of any other kind), CONTRAOCTAVE has to do with infrasonic signalling.

DNS666 July 11, 2013 8:28 AM

@kingsnake: Other rogue AIs sadly missing in action are HAL 9000, GLaDOS, SHODAN, ARIA and AM.

wiredog July 11, 2013 11:36 AM

In many cases the codename (or acronym) is public, and even the type of operation, but the data is classified. Especially if it is an interagency program. DaLAS, for example, is the system that stores all the data from seized digital media, including laptops imaged by US Customs at border crossings. Nothing about the operation of the system itself is classified, just the data.

It stores every file uniquely on a file system in a Big Room of Servers, with the md5 hash and other metadata in a database. That way if a duplicate file is loaded, only the md5 is stored. This also automates making links between persons.

Eric July 11, 2013 11:52 AM

It seems rather peculiar that anyone would put specific project codenames on LinkedIn to begin with. Not only does it risk exposing the existence of a classified project and personnel associated therewith on an open-source (intelligence, not software) resource, but it seems to serve little purpose other than ego stroking. How many hiring managers are going to be searching for someone by “NUCLEON” rather than by, say, “voice to text” and “database analysis”?

MikeA July 11, 2013 12:12 PM

So, who’s working on MD5 collisions to mess with DaLAS?

And as for “Hiring Mangers” looking for NUCLEON, in my experience, HR pays very little attention to what hiring managers asked for, and is not only keyword-based, but requires resumes to be submitted in MSWord format. Yes, even for Linux kernel developer openings. It is impossible to underestimate the intelligence of the hiring process.

Matthew X. Economou July 11, 2013 4:06 PM

Hopefully, no one familiar with MAGINOT BLUE STARS shows up on LinkedIn…

Nick P July 11, 2013 6:35 PM

@ SB

Thanks for the leak. So, the relevant portions would be:

“7. In an unclassified manner, describe projects on which you have worked without citing project names. Unless
you are targeting a specific employer who has knowledge of classified projects, equipment or operations, including covernames in your resume may be counterproductive. Although most covernames are unclassified when there are no associated details or descriptions, they have no meaning to people outside the cryptologic community. It may be better in some instances to avoid the use of covernames and instead, when possible, use unclassified descriptions for projects, equipment and operations.”

This seems to OK it if the codewords are permissible.

@ Economou and Eric


Clive Robinson July 11, 2013 6:36 PM

@ Eric,

    How many hiring managers are going to be searching for someone by “NUCLEON” rather than by, say, “voice to text” and “database analysis”?

It depends, a simple question for you,

    Do you call it a hoover or a vacuum cleaner?

That is some words or names develop a meaning by them selves by frequent usage.

On the odd occasion it’s the same for some codeword projects, which develop a life of their own and become a recognised shorthand for a particular class of activity, one such is TEMPEST.

But yes in most cases the project names remain “confidential” or “restricted” if not “secret”, but over clasifing them makes some tasks such as administration and budgeting difficult when contractors are involved. And this is where the fun starts, if multiple contractors are involved then project names did use to make it onto C.V.s in much the same way as security clearances.

Other nations have different rules that also change from time to time. For instance in the UK back in Maggie Thatchers day, putting your security clearance on a CV would lose you your clearance as would leaving a job. That is the clearance was based on the job not the person and was thus non transferable (in theory). Part of this was done on the simple idea that each organisation you worked for would get you cleared from scratch to prevent certain types of security problems which I gather are currently being investigated in the U.S. This system is known to have caught atleast one “minister level” person out and it cost them their career (investigators found questionable material in the persons library at there home that indicated a prediliction that at the time was not allowed).

The system certainly caused problems for people that might have in effect two jobs requiring clearance. I’m aware of somebodies clearance not coming through because the new request for clearance was at a lower level than the person held for their existing job.

In the UK it got so bad that applying for a “civil service” position had become so drawn out that entry level candidates were finding other jobs rather than wait six to nine months in limbo. And this has apparently given rise to a relaxation of the rules for certain types of candidate…

Nathanael July 12, 2013 1:47 AM

The NSA really doesn’t understand the basic security principle of having the minimum number of secrets, do they? Too many secrets and you have guaranteed insecurity.

Autolykos July 12, 2013 3:42 AM

Couldn’t that be an example of “soft leaking”, i.e. people who like what Snowden did but don’t have the balls to go all the way, and stick with stuff that’s legal to publish and/or looks like a mistake they’re unlikely to get in much trouble for.
@Dave: You mean the guys preparing for SEALION? Haven’t seen those in a while – they might have even axed their budget by now.

Sam July 12, 2013 4:17 AM

Are NSA code names meant to be so dull so as to make them seemingly uninteresting, or are they just bad at picking code names ?

SWAK July 12, 2013 5:55 AM


Some programs are given codes based on the random choice of two words. That why you get names like dish-fire and fast-scope. Other names appear to relate to the content of the program, which is not as secure. Still others appear to be one random word, e.g. fascia.

Analysts can probably tell something about the agency or department that programs come from by the method of naming the program.

Clive Robinson July 12, 2013 6:10 AM


    Some programs are given codes based on the random choice of two words That’s why you get names like dish-fire and fast-scope.

Scott Adams did a Dilbert cartoon on this. He had Catbert as a consultant explaining that they used a list of astronomy words for the first and science words for the second. He then went on to say that, that was the reason the project was called Uranus-Hertz.

If anybody knows the online link can they post it 😉

Michael Edwards July 12, 2013 11:34 AM

Why do I get the feeling that WEBCANDID is a DB of interesting Chat Roulette sessions?

Jim July 12, 2013 1:28 PM

Sam • July 12, 2013 4:17 AM
Are NSA code names meant to be so dull so as to make them seemingly uninteresting, or are they just bad at picking code names ?

Really, wouldn’t their work be super boring? Even for those that “get to” (GHCQ) read anybody’s internet traffic — that is a pretty sad reality show.

It isn’t like they work at Torchwood or in X-Files. There isn’t going to be any memoirs because their lives are too boring. No movies, no books.

There’s probably a gaming effect as it is with people and their jobs. Gaming like, you see a kid playing a video game and they are really into it, but you really aren’t interested in it.

For all the ocean of classified data released [by whatever means], extremely little of it is even interesting in anyway whatsoever.

The sad reality is, for all the spies and spying, for all the behemoth size of the world’s secrecy cultures… truth is, unless you are a direct competitor who actually has competitive concerns, it is completely useless information.

Maybe Zero Dark Thirty was really disinformation, and it wasn’t just disparate facts and lone agents that caught Bin Laden – -but the big data and secrecy machine. (Sarcasm.)

Do they even know why they are in Iraq and Afghanistan?

If they do not even know that for sure, they hardly know anything else important.

Probably one reason why people like to fantasize about Roswell having alien secrets. It all means something!

Lio July 13, 2013 1:59 PM

I’m surprised I can’t find COASTLINE in a quick Google search, much less on It’s a simple tool for sending/receiving KLIEGLIGHT reports.

Leave a comment


Allowed HTML <a href="URL"> • <em> <cite> <i> • <strong> <b> • <sub> <sup> • <ul> <ol> <li> • <blockquote> <pre> Markdown Extra syntax via

Sidebar photo of Bruce Schneier by Joe MacInnis.