No, U.S. Intelligence Has Not Concluded Russian Bounty Story Was a ‘Hoax’ or ‘Fake’

Ramil Sitdikov/Getty Images

Defenders of former President Donald Trump and critics of President Joe Biden have seized on revelations about a U.S. intelligence assessment of the Russian bounty story to claim that it was a hoax or a fake all along.

On Thursday, the Biden administration revealed that “The United States intelligence community assesses, with low to moderate confidence, that Russian intelligence officers sought to encourage Taliban attacks against U.S. and coalition personnel in Afghanistan in 2019 and perhaps earlier, including through financial incentives and compensation.”

When these reports surfaced in the American media last year, they prompted a blizzard of criticism for Trump and his administration because the reporting alleged that Trump was briefed on the bounties and had done nothing. Trump denied he’d been briefed, but reports continued to surface that he had been.

White House Press Secretary Jen Psaki was grilled about the latest revelation, and was even asked if President Biden “regrets” attacking Trump over the reports during the presidential campaign. She did not think so.

The reporting on this assessment is based on a background briefing by senior administration officials on a new raft of Russia sanctions, only portions of which have made their way into the reporting. But in order to properly contextualize the information, it’s important to look at what, exactly, these officials said. Here are the relevant portions of that briefing:

Finally, let me offer an update on reports of bounties on U.S. soldiers in Afghanistan. The United States intelligence community assesses, with low to moderate confidence, that Russian intelligence officers sought to encourage Taliban attacks against U.S. and coalition personnel in Afghanistan in 2019 and perhaps earlier, including through financial incentives and compensation. U.S. intelligence community agencies have low to moderate confidence in this judgment, in part because it relies on detainee reporting and due to the challenging operating environment in Afghanistan.

Our conclusion is based on information and evidence of connections between criminal agents in Afghanistan and elements of the Russian government. This information puts a burden on the Russian government to explain its actions and take steps to address this disturbing pattern of behavior.

…

Q Hey. Thank you. Question — a couple questions. Why list the reported Taliban attacks if there’s low to moderate confidence?

…

SENIOR ADMINISTRATION OFFICIAL: Thanks. I’ll take the first and third questions, and I will hand it over to [senior administration official] for the second one.

With respect to the question of the reports of bounties or financial compensation on — you know, provided to Afghans to kill Americans or coalition troops, as I noted in my opening comments, the IC has low to moderate confidence in this, both because it’s based in part on detainee reporting and because of the difficult operating environment in Afghanistan.

The actions that we have announced today are in response to the cyber intrusions and the election interference. We have noted our conclusion of the review that we conducted on the bounties issue and we have conveyed — through diplomatic, intelligence, and military channels — strong, direct messages on this issue, but we are not specifically tying the actions we are taking today to that — to that matter. We are tying it to the SolarWinds and election interference matters.

I did want to report to you, however, our finding on this, which is: There is an assessment that Russian intelligence officers did seek to encourage Taliban attacks against U.S. and coalition personnel, including through financial incentives and compensation, but because of the low- to moderate-confidence element of this, our focus is on sending a clear message to Russia about the steps the United States would take in response to such behavior were it to continue.

Reporting on the assessment featured characterizations like The Daily Beast’s explainer, which translated the administration’s “low to moderate confidence” in the underlying intelligence thusly: “Translated from the jargon of spyworld, that means the intelligence agencies have found the story is, at best, unproven—and possibly untrue.”

But that’s not really an accurate translation of the way the U.S. intelligence committee assesses confidence levels. Any assessment — high, moderate, or low confidence — would fit that definition. Here’s how the National Intelligence Council defines the various confidence levels:

We ascribe high, moderate, or low confidence to assessments:

• High confidence generally indicates that judgments are based on sound analytic argumentation and high-quality consistent reporting from multiple sources, including clandestinely obtained documents, clandestine and open source reporting, and in-depth expertise; it also indicates that we have few intelligence gaps, have few assumptions underlying the analytic line, have found potential for deception to be low, and have examined long-standing analytic judgements held by the IC and considered alternatives. For most intelligence topics, it will not be appropriate to claim high confidence for judgements that forecast out a number of years. High confidence in a judgment does not imply that the assessment is a fact or a certainty; such judgments might be wrong even though we have a higher degree of certainty that they are accurate.

• Moderate confidence generally means that the information is credibly sourced and plausible but not of sufficient quality or corroborated sufficiently to warrant a higher level of confidence. There may, for example, be information that cuts in a different direction. We have in-depth expertise on the topic, but we may acknowledge assumptions that underlie our analysis and some information gaps; there may be minor analytic differences within the IC, as well as moderate potential for deception.

• Low confidence generally means that the information’s credibility and/or plausibility is uncertain; that the information is fragmented, dated, or poorly corroborated; or that reliability of the sources is questionable. There may be analytic differences within the IC, several significant information gaps, high potential for deception or numerous assumptions that must be made to draw analytic conclusions. In the case of low confidence, we are forced to use current data to project out in time, making a higher level of confidence impossible.

A close reading of the Biden administration officials’ statements would appear to indicate that the low-confidence intel came from detainees, while the moderate confidence intel likely relates to “information and evidence of connections between criminal agents in Afghanistan and elements of the Russian government.”

As the call indicates, these officials assert that the information was strong enough for inclusion in diplomatic communications and warnings to the Russians.

What hasn’t been answered definitively, and doesn’t appear to have been asked, is whether Trump was briefed on the information. However, the revelation is potentially consistent with Trump’s explanation that he wasn’t briefed because the information wasn’t “credible.”