The mosaic concept is the name John Kiriakou gives to the intelligence practice of building a complete picture out of fragments that are individually worthless: “it’s called the mosaic concept, where you have just a little tile, a little piece of information, but then you put it all together at Mossad headquarters and you come up with an entire picture.”[1]
He attributes the idea itself to Soviet doctrine: “it’s something that the Soviets came up with in the 1960s, where they believed that if you look at enough unclassified or declassified information you can put pieces of this mosaic together — or puzzle together — to come up with a classified program.” It is, he adds, “something that the United States has long struggled with.”[2]
As an operation
The clearest modern demonstration Kiriakou offers is Mossad’s networks in Iran. The Afghan refugees recruited into them were not asked for secrets. They were asked where the nuclear scientists lived, where the generals lived, and what time the generals left their apartment buildings in the morning — “all these little” items, each meaningless on its own.[1] Assembled centrally, Kiriakou says, they are what “made Israeli targeting so much easier.”[1]
The implication he draws is about cost rather than sophistication: an intelligence service willing to run enough low-level sources at fifty or a hundred dollars a month can obtain, in aggregate, what no single recruited source would ever hold.[3][1]
As a prosecution theory
The same logic runs the other way, and Kiriakou treats that as a live danger rather than a curiosity. Told that academic researchers had been warned by former government lawyers — including a onetime general counsel of the NSA — that analysing declassified documents could expose them to Espionage Act charges, his response was “he’s right. It’s called the mosaic concept.”[4][2]
He regards the prosecution as realistic rather than theoretical: given a prosecutor in the Southern District of New York, “or more likely in the Eastern District of Virginia, who wanted to make a name for himself — sure, I could see a professor being charged with espionage for looking at unclassified or declassified information. Absolutely.”[5]
The doctrine also supplies the official justification for over-classification. The reason officers are trained to stamp even trivial material, Kiriakou says, is the aggregation argument: leave it unclassified and “some malignant force out there — the Russians, the Chinese, whomever they might be — might be able to put it together and come up with some classified information as a result.”[6]