Thread by @michaeldweiss

[Thread.] Few comments about this NBC story, which is yielding the usual feast of reason and flow of the soul on social media. nbcnews.com/politics/natio…

The piece is centered on comments made by Gen. McKenzie, who, from what he says and how the story is framed, wants to see a straight line from GRU payments to attacks on U.S. forces in Afghanistan.

The term "bounty" was never used in the intelligence, as per @douglaslondon5 and others in the IC, although it became the media's term of art following the first NYT story.

But what was the intelligence? The NYT was a lot more specific than NBC. The money, $500K, came from GRU Unit 29155, which controlled bank accounts that transferred the funds to Taliban-linked accounts:

A lot of earlier speculation, that the intel was premised solely on what captured and interrogated Taliban fighters said, was wrong. The gravamen, which even convinced IC skeptics that the GRU was incentivizing attacks on Americans, was intercepted financial data, as per NYT:

Now why did these intercepts raise alarm bells at CIA? Because of what Unit 29155 is within the GRU. They're an assassination and sabotage squad responsible for the Skripal and Gebrev poisonings, a failed coup in Montenegro and other violent acts of destabilization in Europe:

(It would be one thing, for instance, if a Russian service was simply recruiting Taliban commanders for espionage: to garner intel on NATO movements in Afghanistan or ISIS activity or the opium export industry, etc. But Unit 29155 doesn't do that.)

Now what, based on the reporting, is in dispute among U.S. intelligence officers and the Pentagon? That Russia is paying the Taliban and has been for years? Nope.

That the GRU has been paying the Taliban? I've seen no "debunking" of the original NYT scoop suggesting that is in doubt among skeptics. That Unit 29155 is the culprit? Ditto. But there are hints and clues. E.g.:

One way to stand up the "bounty" assessment is to draw a straight line between prior attacks (particularly lethal ones) on U.S. forces in Afghanistan and the money tracked going from Unit 29155's account to the Taliban. This would constitute evidence that the program not only...

... exists but has been successful. McKenzie, it convincingly appears to be the case, is trying to do just that. Here again is his quote from the NBC story, followed immediately by another "U.S. military official familiar with the intelligence" who says:

McKenzie's "final connection" would be American corpses -- or at least documented Taliban attacks -- with Unit 29155's paw prints on them.

But herein lies the problem, which gets at the broader ambiguities in intelligence gathering and intelligence assessing. Operational intent is different from operational success or even operational action.

It may well be the case that, owing to the newness of this alleged "bounty" program (the NYT dated it as beginning in early 2020, though other outlets dated it even earlier), none of the Taliban fighters who received the money ever went out and tried to kill any U.S. soldiers.

Maybe they were rounded up or got croaked before they could, maybe they got cold feet, maybe they took the money and fucked off.

So you've got generals, who want to know whether or not to retaliate for Russian aggression, looking for copper-bottom "actionable" proof, while spies are making interpolations based on context (Unit 29155's m.o.) and available evidence (intercepts, detainee testimonies).

And all of the above assumes you're dealing with professionals who are eager to suss out the truth and share a baseline consensus on evidence but disagree over conclusions. It doesn't address how intel can be embellished or obfuscated for political purposes.