The problem of military intelligence being given away by social networking is a very old one. Directly useful OSINT (i.e., extranet social postings directly translatable to location) has become a rarity among contemporary military forces, due to greater technical capabilities and COMSEC discipline; however, negative social OSINT is still an issue. When extranet social postings from military-affiliated or -contracted individuals cease due to COMSEC, it is possible to infer that this is due to a forward deployment, and for a competent MILINT analyst to deduce from this such information as units deployed, specializations involved, possible deployment locations, etc., etc.
To address this issue, the Fifth Lord of the Admiralty created NATTER SPATTER.
Under the aegis of OPERATION NATTER SPATTER, the Stratarchy of Data Warfare maintains a team of AI forgers which monitor the extranet social postings of Imperial Military Service personnel and contractees, learn how to precisely model and imitate them, and are prepared to step in at a moment’s notice when COMSEC takes effect with a consistent stream of forged data – memeweave postings, imagery, slinky recordings, and even exomemories – indirected in such a manner as to be indistinguishable from the genuine article, and demonstrating a continued civilian life. A stochastic social event generator seasons the basic event stream with sufficient nonrepresentative variation to prevent NATTER SPATTER content from being identified by excessively consistent mundanity.
While NATTER SPATTER postings will not, naturally, survive a full consistency check against the complete event record, public and private, such a full consistency check is beyond the capacity of many intelligence agencies and serves, in other cases, to drive the cost of obtaining this type of OSINT higher in terms of computation and time than can be justified.
“We think the Imperial Navy is going to move against us.”
“Why’s that?”
“MODESTY BLAZE just started posting the really sappy, romantic memes in the last download cycle. Major uptick in those memes.”
“How big?”
“Posted a two yobibyte cute cat video of ginger cats.”
“Oh (BEEP), we’re (BEEP)ed, aren’t we?”
On the first read, that seems very creepy: they’re forging interactions from someone and that seems like fraud, in Imperial culture.
On reflection, a lot of that could be mitigated if the soldiers involved were in on it. In which case it could lean more towards trolling, be more consensual, and a chunk of the seeding could well be provided by the sophants in question after they’ve been checked and run through whatever censors are being used. “Meme running wild through the office” and “meme running wild through the base” are close cousins, after all.
…does it count as consent if the person who consented then had their memories of consenting (presumably temporarily, and probably consensually, though then there might be awkwardness about those memories…) edited out for infosec purposes?
But yeah, in the absence of active consent that does sound a lot like it could qualify as fraud… obviously assuming it gets noticed first, which would be tricky at best.
Oh, to be clear, the IMS members involved, their families, and hell, even their close friends are all read in on NATTER SPATTER.
(In classification level terms, its existence is RESTRICTED INFRARED, which is the step below CONFIDENTIAL/SENSITIVE and immediately above “it’s not classified but for gods’ sake don’t post this to the War Thunder forums”.
Technical methods are rather more classified, of course, and that COMSEC is active is as secure as the operation invoking the COMSEC, but the existence of the program isn’t much of a secret, on the grounds that knowing about the existence of the program doesn’t actually help the oppo, and may actually hurt them inasmuch as it might encourage them to spend lots of time, money, and compute on trying to break it.)
I also haven’t decided on it, yet, but I’m considering the possibility that they go back and flag all the NATTER SPATTER posts once COMSEC goes off.
Valid on the memories-related part, I guess. Hadn’t thought of it quite that way…
The uncomfortably-close-to-fraud situation I was thinking of was more something to the effect of “I did not, in fact, have a dinner at the Three Bandals*, on account of not actually being particularly interested in that kind of cuisine, and I know that it’s your stochastic generator that suggested it but I would also really appreciate if people who would look me up later did not incorrectly assume that I’m a fan of that place”.
But of course anything like that would happen (if at all) well after any COMSEC had already gone off.
*) I was unable to think of a plausible name for an Imperial tavern that did not make it sound like a high-class restaurant