Over the July 4 weekend, the New York Times reported that Director of National Intelligence (DNI) John Ratcliffe has produced a memorandum that, on the one hand, acknowledges the assessment by the CIA and National Counterterrorism Center that Russia paid bounties to the Taliban to kill American service members but, on the other hand, may seek to cast doubt on this assessment to serve the White House’s political purposes. According to the Times, concerns about politics infecting the process stem from the timing and the reported content of the three-page Sense of the Community Memorandum (SOCM), a product of the National Intelligence Council (NIC), which reports directly to the DNI. The Times also quotes former intelligence officials who express concerns about the potential politics at play. One of those former officials served as chair of the NIC, and another was the predecessor to one of us as director of the National Counterterrorism Center. Based on what has been reported by the Times and other outlets, we share their concerns.
That’s a painful conclusion to put forward as it would mean the country’s top intelligence official is manipulating intelligence processes for political purposes in the context of a direct threat to American service members’ lives. That we even have to ask the question of whether this intelligence is being politicized is a measure of how far from normal business we have strayed in recent years.
So, why the concern over the SOCM product? First, we take no issue with the likely authors and reviewers at the NIC. One can come to the conclusion that the SOCM was tasked for a largely political purpose while, at the same time, believing that it was prepared in a professional and honest manner by analysts at the NIC. At least we have no reason to believe otherwise, and if its conclusions were significantly edited or shaped by DNI Ratcliffe, the public would likely learn of that soon. That crude form of manipulation would strike at the heart of analytic integrity, and it would probably leak if Ratcliffe or his staff at DNI inserted themselves too forcefully in the drafting or editing process.
That said, even a completely “professional” product can still be highly susceptible to political manipulation—and there are strong reasons to believe that’s what has happened here in light of all of the reporting that we have to date, including statements by lawmakers who have been briefed on the subject. How would this form of political influence work? An intelligence product that presents a relatively firm bottom line conclusion that the intelligence community believes this Russian activity was in fact occurring could still be cherry-picked to suit a political purpose and drive a particular political narrative. “NSA doesn’t agree,” “There’s no real consensus,” “The intelligence is muddled and unclear,” “The agencies are all over the map,” “There are intelligence gaps that need to be filled before we can trust the conclusions.” All of these are characterizations of aspects of the memo that could be technically true but would still make the product susceptible to political manipulation.
In particular, each of these narratives could be drawn from a professionally and honestly produced analytical work and would obviously suit the White House’s desire to portray that the intelligence wasn’t ready for prime time, that there was no need to brief the President, and that this was hardly enough to drive action to engage Putin.
And that narrative would basically be a form of propaganda that would make Soviets blush. That’s because the gaps and discrepancies within the underlying intelligence — if accurately reported, including those contained in the SOCM — are basically an everyday occurrence when it comes to the process of analyzing sensitive intelligence on important national security topics and presenting policymakers with the information they need to make decisions and pursue courses of action. In other words, the characterizations above could be, in a narrow sense, accurate; but the takeaways that the White House intends to draw and disseminate from them could still be wildly misleading.
The timing of the SOCM product — apparently tasked only after the story first broke publicly and then produced within a matter of days — also calls into question the motivations of DNI Ratcliffe. If there was a real hunger among policymakers at the White House to understand better the underlying intelligence and the nuanced (and possibly competing) views of analysts across the Intelligence Community, we suspect that tasking would have been issued months ago—when the intelligence on life-or-death threats to U.S. service members first surfaced. On the other hand, if the SOCM was tasked only lately as the Times reporting suggests, one can’t help but wonder if the underlying purpose was to give the administration ammunition to cast doubt on the intelligence community’s bottom-line conclusion. All told, the recent timing of the memo’s tasking and rush to complete it suggests it was intended not to inform policy discussions on how to protect American troops but to inform political efforts to rebut the media reports and bipartisan concerns on the Hill.
What’s more, if the underlying intent was political, one can imagine the DNI’s office might have tasked the memo in such a way as to elicit some of the answers that would be useful to a White House that wants to argue that this is all overblown. “Is there complete Community consensus on the conclusions in the PDB back in February?” “What are the varying levels of confidence in the assessment presented?” “Does the IC have direct evidence as a basis for reaching a judgment with high confidence?” “Does the NSA have signals intelligence that provide the agency a basis for its assessment?” “Does the IC have information that connects President Putin or other Kremlin officials to the operation?” Answers to these kinds of specific questions might well have provided the Administration with just the cherry-picked characterizations that they would need to bolster their argument that there’s nothing to see here—even though the answer to any of these narrow questions would not actually undercut the idea that there’s a serious threats to American service members to which the Trump administration has failed to respond.
Indeed, reports on the content of the SOCM suggest manipulation and skewing of its conclusions. Consider three specific items in the text of the two-and-a-half-page memo as reported.
First, the SOCM “stressed that the government lacks direct evidence of what the criminal network leaders and G.R.U. officials said at face-to-face meetings,” according to the Times. The absence of direct evidence of those meetings is not actually a surprising or even notable gap in the context of assessing a Russian covert operation, given two factors: the unlikely nature of the United States gaining intelligence on the exact words spoken between Russian and Taliban officials in secret meetings, and the mosaic of other information the IC has reportedly obtained indicating that the bounty arrangement existed. So why stress its absence? If the Russians are actually engaged in the reported behavior in Afghanistan, they have every incentive to make their activities as deniable as possible, including by hiding their fingerprints, using proxies where they could, and denying at every turn. Yet it’s easy to imagine Ratcliffe and other Administration surrogates in the coming days describing the memo to make this intelligence gap sound like a bigger hole than it really is.
Second, the Times reports that the National Security Agency “did not have information to support that conclusion [of the Russian bounty operation] at the same level [as CIA and NCTC], therefore expressing lower confidence in the conclusion.” And the Times goes on to say that the SOCM then “emphasized that the National Security Agency did not have surveillance that confirmed what the captured detainees told interrogators about bounties.” Once again, the absence of one specific type of intelligence—signals intelligence—is unsurprising here, given that it’s unclear how the Russians and Taliban are communicating about the bounty arrangement and given that there are (reportedly) other kinds of intelligence, from detainee interrogations to financial transactions, that indicate the arrangement exists. So why would the SOCM emphasize this absence? What’s more, the question to be asked of NSA by the NIC is whether it agrees with the all-source judgment on the existence of the operation, not whether the agency has its own information to support the conclusion at the same level of confidence. We hope that’s what NSA’s input reflected. It is notable that the Wall Street Journal reported last week that “the differences [between the NSA and CIA] weren’t over the central assessment that operatives with Russia’s GRU intelligence agency paid bounties to the insurgent Taliban movement to kill Americans,” according to some of the people familiar with the matter. That suggests the SOCM may have been produced in a way that amplifies less meaningful differences between the agencies.
Third, the Times reports that the memo identifies gaps in intelligence on the issue of specific attribution for the bounty operation within the Kremlin. “The memo also says that the Defense Intelligence Agency did not have information directly connecting the suspected operation to the Kremlin,” according to the Times. But when it comes to a GRU unit acting inside Afghanistan like this, it may be exceedingly difficult to directly connect it to Putin or other senior Kremlin officials. That’s of no consequence for what’s at stake here. Indeed, if the IC has compelling information that the GRU section was placing bounties on American soldiers, that’s more than ample information to press Putin to have it stop. Meanwhile, having this question of attribution raised in the SOCM is all too convenient for a White House interested in alluding in abstract terms to purported gaps in the intelligence.
In order to understand why the SOCM product came to look the way it did, at least as reported in the Times, it’s important to understand how the analysis probably emerged, with nuanced and caveated conclusions and a lack of complete consensus among the expert analysts. That in fact is the norm in intelligence work. It is rarely if ever the case that the United States or any other country’s agencies would have a singular gold nugget of incontrovertible intelligence that would “prove” the case on something like this. For example, it’s the stuff of fiction, not real intelligence work, to expect the United States to intercept Putin’s personal aide saying to the GRU chief: “The Boss has green lighted the bounty program in Afghanistan. Go for it starting on Tuesday.”
Instead, IC analysts looking at this material would have been trying to piece together a complicated mosaic, with information of all different kinds and differing degrees of reliability. There might have been one triggering piece of compelling intelligence from one part of the intelligence community, but that would have led to a careful effort to identify and evaluate other intelligence to corroborate what that initial reporting suggested. Analysts would also have been putting the information into the context of what they know from long observation about Russian interests; about Putin’s practices, including ratcheting up support for the Taliban over time; about the GRU’s connections to Afghanistan; about the Taliban operatives; and other information. Russia analysts would have been involved, Afghan-focused analysts would have been involved, terrorism/counterterrorism analysts would have been involved—all bringing their special expertise and perspective to bear. The analytic process would have been grinding hard on information of this sort over weeks and months.
We’ve painstakingly described the possible misuse of the SOCM and the underlying analytic process that probably led to its reported conclusions because those aspects of this story might help us understand why the administration seems so reluctant to credit the seemingly obvious intelligence conclusion that Russia is acting against our interests in Afghanistan. And that’s the real story here. The inside baseball on intelligence work is not the part of this story that matters most.
Why do we say that? Because intelligence on “hard” targets like Russia or North Korea or Iran rarely comes packaged in the single “gold nugget” form that we alluded to above. And if any intelligence picture that falls short of meeting that standard couldn’t be shared upward to the President, or if it couldn’t serve as the basis for considering concrete action to respond to credible threats to American lives, then our foreign policy would be truly paralyzed. We would simply never meet the threshold to take action, let alone to consider it. So it’s simply not the case that we can’t believe our eyes unless we have the single golden nugget of intelligence.
For all of the complexity of the intelligence process discussed above, the basic facts are clear: Our intelligence community has assessed that Russia is acting in a way that threatens American lives in Afghanistan. Yes, there are nuances and varying degrees of confidence in that conclusion. That’s normal. But there should not be any kind of debate about whether this conclusion should have been presented to senior policymakers and the President for their review and action—nor any doubt that they should have responded, urgently. In any other time, that would be a given. There might be a policy conversation to be had about what exactly to do in response, but no reasonable conversation can be had about whether to put this conclusion on the table at the level of the President and his National Security Council for discussion about what actions to take.