In Russian bounty debate, once again this administration lacks intelligence
© Getty Images

Let’s lay aside for a moment the question about whether President TrumpDonald TrumpHead of firms that pushed 'Italygate' theory falsely claimed VA mansion was her home: report Centrists gain foothold in infrastructure talks; cyber attacks at center of Biden-Putin meeting VA moving to cover gender affirmation surgery through department health care MORE was or wasn’t informed about the U.S. intelligence assessment that Russian military intelligence — the GRU — paid bounties to Taliban fighters in Afghanistan to kill Americans. As a former career intelligence officer who helped produce the President’s Daily Brief, it’s hard to envision any plausible scenario in which a president should not be briefed on such a matter. But this president, as we know, is different.

I’ve spoken with former White House officials who confirm that the president does not like to hear information that is inconvenient or uncomfortable for him. So perhaps he really wasn’t told about this intelligence suggesting Russian complicity in the deaths of at least some of the 20 Americans killed in Afghanistan in the past year.

Most disturbing for me is the argument from senior administration officials that President Trump should not have been briefed on these latest Russian allegations because the intelligence was “not conclusive.” We hear this in various guises from national security advisor Robert O’Brien and from new Director of National Intelligence John RatcliffeJohn Lee RatcliffeCentrists gain foothold in infrastructure talks; cyber attacks at center of Biden-Putin meeting Five things to know about the new spotlight on UFOs Extraordinary explanations for UFOs look increasingly plausible MORE.


Meanwhile, a White House spokesman indicated last week that the intelligence did not go to the president because there was some dissent about its veracity. These officials apparently think that intelligence, before it goes to the president, must be substantiated beyond doubt. To use a cliché, they want “smoking guns.” Intelligence does not work that way.

We’ve seen this misunderstanding before in this administration, an apparently willful ignorance of intelligence and how it operates.

Consider the October 2018 murder of U.S.-based journalist Jamal Khashoggi at the Saudi Consulate in Istanbul. The CIA’s analysis in February 2019 concluded that Saudi Crown Prince Mohammed bin Salman probably ordered the killing. However, Secretary of State Mike PompeoMike PompeoWhite House denies pausing military aid package to Ukraine The Hill's Morning Report - ObamaCare here to stay The Hill's Morning Report - After high-stakes Biden-Putin summit, what now? MORE — a former CIA director, no less — issued a sophomoric dismissal of the Agency’s assessment, saying that there was “no direct reporting” linking the murder to the crown prince. Again, the elusive search for the “smoking gun,” which betrays a dismaying denial of what intelligence assessments are all about and how intelligence analysts arrive at them.

Someone very brave and senior in the U.S. Intelligence Community needs to inform these senior Trump officials, “We have intelligence when we don’t have a smoking gun.”

Intelligence by necessity makes judgments about what is not known conclusively, because “smoking guns” are rare. Almost all intelligence questions have two or more sides to them. Information is often contradictory. Intelligence analysts try to sort it all out, not aiming for an ironclad, unassailable conclusion (though that does happen, rarely), but to arrive at what has been called “best truth”—the most complete assessment of the situation that the evidence supports.


What is missing from the debate over the Russian GRU bounties is the key question of analytic confidence. In the aftermath of the intelligence failure over Iraqi weapons of mass destruction more than 15 years ago, the Intelligence Community instituted a process of asking analysts how confident they were in their own judgments, based on the intelligence sources available. It is now routine for key intelligence assessments to be accompanied with expressions of “confidence levels.”

What do these confidence levels mean? A rare explanation comes from former CIA director Michael Hayden, who in 2007 briefed the White House about a suspected Syrian nuclear reactor. CIA at the time made “high confidence” judgments that the Syrians had a nuclear reactor and that they were involved in nuclear cooperation with North Korea for years because these things were directly observable. The judgment that the North Koreans had built the reactor was made with only “medium confidence” (despite it being exclusively a North Korean design) because the CIA had no “eyes on” Koreans actually constructing the site.

In other words, intelligence can be good — even very good — without its being “conclusive.”

If news reports are true, CIA made the 2019 assessment about the Khashoggi murder with “medium to high confidence,” suggesting the evidence as a whole leaves little doubt that the Saudi crown prince directed the murder, even though U.S. intelligence did not collect the actual order. In the current debate, a key question is how confident the analysts are in the intelligence judgment that the Russian GRU paid bounties for the murder of Americans.

Trump administration officials appear to be playing politics with intelligence, which should remain apolitical. They — and the president himself — do this country a great disservice by dismissing intelligence when it is not 100 percent proven, conclusive or vetted and held to some unrealistic standard.

For the sake of our troops and for national security generally, they need to be less political and more intelligent about intelligence.

Nicholas Dujmovic directs the intelligence studies program at The Catholic University of America in Washington, D.C. He is a retired CIA officer and served as an analyst, manager, and staff historian.