Evaluating the Buzzfeed dossier, by a former Intelligence Analyst

Jim Arkedis on 2017-01-12

UPDATE on February 13, 2017: I published an update to this story here, following CNN’s February 10th report that the intelligence community had confirmed some aspects of this dossier.

Original article:

Over the last 48 hours — since Buzzfeed published an intelligence dossier alleging the Russian government’s involvement with President-Elect Trump — few words have been subject to as much debate and scrutiny as “unverified.” The entire dossier is indeed unverified but that does not mean it contains no truth.

Further, as a former intelligence analyst for the Department of Defense, I can tell you that “unverified” does not mean fake. There is much we can glean from within the report by assessing its many individual sources based on how the report’s author presents them.

I can also tell you that just because the Justice Department determines a report does not meet the higher legal standard of proof (IE, “beyond reasonable doubt”) does not mean it lacks intelligence value.

Of course, the intelligence world does not accept every unverified report as true. Quite the opposite.

As an intelligence analyst, my job was largely to examine the many unverified pieces of information that the Intelligence Community collects to determine the likelihood that a given report might be true. Given the nature of raw intelligence, it’s often impossible to verify a report’s truth beyond a shadow of doubt. It’s also important to remember that intelligence analysts aren’t in that business. The intelligence community is in the business of identifying “credible threats” to the safety and security of the American people.

Assessing “Credibility”

Below is how I would assess the credibility of the sources and allegations detailed in Buzzfeed’s recently-released dossier and an explanation of why I believe its two main allegations should be judged on their individual merits as credible with moderate-to-high confidence.

No, that’s not the same as saying the allegations are 100 percent guaranteed to be true, but I think there’s enough evidence there that it would be irresponsible not to consider how this could impact our nation’s security and what, if anything, can be done to mitigate those potential impacts.

I’m sharing this to shed light on how the intelligence community evaluates sources and reports and why it’s inappropriate to use the Justice Department’s standard of proof.

There is a degree of merit to these allegations and thus a risk to our national interests. The result is that our security organizations and investigative agencies are correct to look in these matters further.

Evaluation of the source of the information, the former MI-6 agent:

According to multiple press reports, the source is Christopher Steele, “former British MI-6 agent with good network in Russia.” This source is very likely credible, for two reasons:

  1. Source has been vetted by current US officials at the CIA and deemed credible. This vetting likely would have established the following points:

2. At least two US political entities believed him credible enough to invest “tens of thousands of dollars.”

It is common for a former intelligence collector to establish a private intelligence collection firm as a way to monetize the collector’s network of sources. The biggest risk with this type of source is that s/he has a financial concern to provide information that pleases his clients.

Next, I evaluate the former MI-6 agent’s sources to substantial individual charges:

Generally, highly credible sources must have placement near the information as a virtue of their position and direct access to the specific information reported. Then we look to assess whether that source’s information can be corroborated by a separate source with placement and access.

From the dossier, it’s important to focus only on two main charges in the dossier:

First, that Donald Trump, his organization, and his campaign operatives have had an extensive relationship with the Kremlin.

Source Evaluation:

Source A is a “Senior Russian Foreign Ministry figure active inside the Kremlin”.

Source B is a “former top level Russian intelligence officer still active inside the Kremlin.”

Source C is a “senior Russian financial official.”

Source D is a “close associate of Trump who has facilitated his travel to Moscow.”

A note about “unknown credibility.” Do not confuse “unknown” with “low credibility” or “not credible”. Unknown means unknown, nothing more, nothing less. While all intelligence should be corroborated, it is especially important in the case of sources of “unknown credibility.”

Accusation: Sources A and B make the claim of an extensive relationship between Trump and the Kremlin. Source A has said Kremlin has been feeding Trump information on Hillary Clinton for years. Source B confirms the Trump operation and has said Putin has personally directed it.

It is reasonable to assume these two pieces of information were collected independently; the MI-6 agent would have understood the importance of doing so. Therefore, this is two-source, independent corroboration by sources assessed as highly and likely credible.

Source C has provided contextual information confirming, and going beyond, Sources A and B. This is further corroboration, a positive development. Source C claimed to have “overheard Putin” talking to associates. This is a substantial claim, and could detract from Source C’s credibility as the source may have been trying to impress the MI-6 agent by bragging about access to Putin. However, given Source C’s main value is in corroborating Sources A and B, this is a relatively minor concern.


Based on the direct access to the information and long track record of two established sources and the corroboration of a third and then fourth source on the broad, general issue of Trump’s five-year relationship with the Kremlin, I would assess as credible with moderate-to-high confidence.

Second, that Trump’s sexually deviant behavior at the Ritz-Carlton subjects him to blackmail.

A contextual point: Catching people on tape in sexually compromising positions for blackmail purposes is a well-known piece of tradecraft uniquely associated with Russian intelligence. It is called a “honey pot.”

Source Evaluation:

Source B: See above.

Source D: See above.

Source E: A Western employee of the hotel.

Source F: A different hotel staffer.

Accusation: Source D, of unknown credibility who had facilitated Trump’s travel to Moscow, says s/he was “present” when Trump engaged in deviant sexual behavior. “Present” is a questionable term, one used by the MI-6 agent that may not actually reflect Source D’s physical presence in the room. Source D says that hotel is known to be bugged by FSB.

Sources E and F, both of unknown credibility, independently confirm the story of Trump’s sexual deviance in the Ritz-Carlton. Source B, the “former Russian intelligence officer who is likely credible,” says Russian intelligence has years worth of embarrassing information on Trump that could subject him to blackmail.


It is reasonable to have doubts about the individual details of Trump’s sexual deviance owing to the unknown credibility of the sources. The former-MI-6 agent obviously shares these doubts, which is why s/he went to the length of establishing three corroborating sources. However, it is more reasonable to lend credibility to the fact that the FSB uses the Ritz-Carlton as a collection location for honey pot operations. Given known Russian trade-craft and Source B’s high-level corroboration of Russian intelligence’s confirmation of a broader dossier of blackmail-able information lends additional credibility to general story of sexual deviance at the Ritz. I would assess that the FSB has compromising material on Trump that would subject him to blackmail with moderate-to-high confidence.

One final note: Individual details, like lawyer Michael Cohen’s trip to Prague or the spelling of a name or two, may indeed be disproven. Not everything in these reports is 100% accurate.

However, it is extremely important to emphasize that micro-level inaccuracies do not detract from the credibility of the two broad points that I establish above: that Trump’s organization has had a relationship with the Kremlin and that he is subject to blackmail.

UPDATE: A second “final note”

In a New York Times article, an investigator for a firm in London commented, “where’s the evidence, guys? I can’t use this.” He continues to say that “there’s no explanation of the credibility of the sources.” On the one hand, the commentator is correct: none of the sources are labeled as “credible” within the dossier, which is why I undertook the task of doing it myself based on what we known and can reasonably infer. The point of my piece is that we can confer a degree of credibility based on three aspects: the vetted credibility of the MI-6 agent, the general descriptions of the sub-source’s placement and access, and cross-source corroboration. It might not be enough to use, say, in a political campaign, but it is enough to warrant further investigation due to the potential impacts on U.S. national security.

About the author: Jim Arkedis was a Defense Department intelligence analyst for the Naval Criminal Investigative Service from 2002–2007, specializing in counter-terrorism and counter-intelligence. From 2007–2012, he ran the Progressive Policy Institute’s National Security Project. From 2013-present, he has the president of 4DPAC, a national security-focused political action committee. Reach him on email and Twitter.