
Inspector General Michael Horowitz’s recent report on Foreign Intelligence Surveillance Act (FISA) abuse excoriated the FBI on many fronts. Ten of the 17 instances of “significant inaccuracies and omissions” from the FISA applications coalesced on confidential human source (CHS) Christopher Steele.
Given that Horowitz also found that Steele’s election reporting “played a central and essential role in the FBI’s and Department’s decision to seek the FISA order,” these mistakes are inexcusable. But when the most recently identified malfeasance is considered in conjunction with the IG’s previous—and much-overlooked—audit of CHS validation, the findings are horrifying.
Much of the press and public missed the release in November 2019 of the IG’s audit of the CHS validation process. At the time, I highlighted four blockbuster details buried in the 63-page report, but now knowing the results of the FISA investigation, two aspects of the IG’s audit surface as exceedingly important.
Hiding Problems with FBI Sources
First, as I noted at the time, the IG’s audit of the CHS validation process revealed that the FBI was “intentionally failing to document confidential sources’ credibility and reliability problems so defense attorneys do not learn of them”! Multiple intelligence analysts told the IG “that they received guidance to only state the facts and not to conduct analysis, report conclusions, and make recommendations in the Significant Source Review Panel validation reports.” The IG report illustrated the effect of this: “One Intelligence Analyst told us that he was permitted to recommend a CHS receive a polygraph or operational test to the handling agent by phone but not permitted to document the recommendation in the CHS’s validation report.”
The IG audit also reported that “multiple FBI officials told us that they believe that field offices do not want negative information documented in a CHS file due to criminal discovery concerns about the CHS’s ability to testify.” In fact, as “one FBI official” told the IG, “some U.S. Attorney’s offices will not use a CHS at trial if there is a negative documentation in the CHS’s file.”
These conclusions change the tenor of Horowitz’s findings of FISA abuse. In isolation the report’s summary of the 17 significant inaccuracies and omissions appears the result of gross incompetence, plus an instance of likely criminal conduct by an attorney who altered an email. (There were actually eighteen significant inaccuracies and omissions, when counting the one Horowitz’s team missed: that the FISA applications inaccurately portrayed Steele’s network of sources as a holdover from his time as a MI6 spook.) The perception of gross FBI incompetence applies equally to the subset of ten errors related to Steele.
A Culture of Corruption at the FBI
The IG’s report on CHSs issued the prior month made clear that the CHS files weren’t squeaky clean because the sources were or because of mistakes, but because FBI agents intentionally excluded derogatory information to maintain the sources’ apparent credibility and reliability. That conclusion from the previously released IG audit on CHSs changes everything.
We are not talking about mere mistakes; we aren’t even talking about recklessness. We are talking about a corrupt culture that buried bad news to bury the bad guys (and a few innocents along the way).
The IG’s report on CHSs confessed this conclusion, but stated the proposition so pedantically it was unclear to lay readers: “By withholding potentially critical information from validation reports, the FBI runs the risk that prosecutors may not have complete and reliable information when a CHS serves as a witness and, thus, may have difficulties complying with their discovery obligations.”
“Their discovery obligations” means “to cough up negative information about a CHS.” And do you know where those obligations come from? The U.S. Constitution.
Horowitz’s more recent report on FISA abuse mentions these earlier findings, but in such an innocuous way as to bury the significance: “The OIG’s Audit Division recently completed a review of the FBI’s CHS validation processes finding, among other things, that FBI validation personnel were discouraged from documenting conclusions from CHS validation reviews in their written reports,” Horowitz’s report read, in a “nothing to see here, move along” tone.
But the IG’s earlier findings that FBI agents were “discouraged” from including derogatory information in a CHS’s file has a huge bearing on Horowitz’s FISA report and analysis. Consider, for instance, the backdrop included in the IG’s FISA report about the use of CHSs in FISA applications.
This Deception Applies to FISA Applications
Horowitz’s report explained that CHS’s information may be used “in FISA applications without revealing the identity of the CHS, so long as the handling agent provides the relevant FBI Headquarters operational unit, . . . with the CHS file number, duration of service to the FBI, and a statement on whether the CHS is reliable and has provided reporting that has been corroborated.”
The report also explained that “all information provided to support a FISA application must also be documented in the CHS’s Delta file.” Delta “is the FBI’s official electronic record-keeping system for CHS management.” The report further said the CHS handling agent must review “the facts presented in the FISA application regarding the CHS’s reliability and background,” and state as part of the Woods procedure the FBI is supposed to follow that “based upon a review of the CHS file, the facts presented in the application concerning the CHS are accurate.”
Of course, that never happened with Steele. Steele’s handling agent did not verify the FISA application’s description of Steele’s reporting and, “according to the handling agent, he would not have approved the representation in the application because only ‘some’ of Steele’s prior reporting had been corroborated—most of it had not and because Steele’s information was never used in a criminal proceeding.”
It isn’t merely a matter of Steele’s handling agent not corroborating the FISA source classification. The agents working Crossfire Hurricane also did not document in Steele’s Delta file many details relevant to his credibility, reliability, and bias. We know from the IG’s previous CHS audit that this isn’t just sloppiness—it is an institutional penchant for prettying up a CHS’s profile by omitting negative information from the source’s file.
This reality casts in a much more sinister light the IG’s conclusion that “there were instances where information we deemed significant about Steele was not included in his Delta file.”
What Information Was Omitted about Steele
What, then, to make of the information omitted from Steele’s Delta file?
Much was omitted. Among other things, the IG slammed the FBI’s Validation Management Unit for excluding in its validation report its conclusion that it had failed to “identify any corroboration for Steele’s election reporting among the information that the Crossfire Hurricane team had collected.”
The IG also found fault in the FBI’s failure to add to Steele’s CHS file that in a January 2017, interview one of Steele’s primary sources made statements raising significant questions “about the reliability of” Steele’s reporting that had been included in the initial FISA application. The IG further noted that Steele’s CHS file omitted details concerning Steele’s “multiple contacts with representatives of Russian oligarchs with connections to Russian Intelligence Services (RIS) and senior Kremlin officials” and concerns raised by the FBI Transnational Organized Crime Intelligence Unit concerning the frequency of these contacts.
The IG’s investigation into the Carter Page FISA applications also faulted the FBI for failing to include information in Steele’s CHS file that agents gathered during a November and December 2016 trip abroad to speak with individuals “who previously had professional contacts with Steele or had knowledge of his work.”
Information gleaned from these individuals included statements that Steele “[d]emonstrates lack of self-awareness, poor judgment,” “pursued people with political risk but no intelligence value,” “didn’t always exercise great judgment,” and it was “not clear what he would have done to validate” his reporting. However, the IG found that none of these negative assessments were “memorialized in Steele’s Delta file and therefore not considered in a validation review conducted by the FBI’s Validation Management Unit.”
Here, we find also an intriguing footnote, dropped after the IG notes that the information from these meetings “was not memorialized in Steele’s Delta file:”

Reading beneath the blackened lines, one might hazard that FBI Counterintelligence Division Assistant Director Bill Priestap, who approved the case opening, “may have made” a commitment to British intelligence agents that he’d keep the negative feedback on Steele on the down low. But you can’t accurately swear out a surveillance application when information is missing from the file!
The FBI Didn’t Verify Secondhand Information
There is a second significant problem bearing on the Page FISA applications that the IG exposed in his earlier audit of the CHS validation process: “The IG’s audit of confidential human sources reveals that the FBI did not maintain validation reports for sub-sources, defined in the report as individuals ‘who directly acquire[] information that is then provided to the FBI by an FBI CHS.’” Sub-sources, of course, are secondhand sources of information to FBI sources.
According to Horowitz, Delta “does not identify and track extraterritorial sub-sources.” That means that the CHS system “will lack complete and accurate information on its CHS coverage stemming from extraterritorial sub-sources.”
The FBI’s failure to maintain a database to identify and track sub-sources decimates the entire basis for the Carter Page surveillance warrants. According to the IG report, the FISA applications all disclosed that “Steele relied upon a Primary Sub-source who used a network of sub-sources, and that neither Steele nor the Primary Sub-source had direct access to the information being reported.” The FISA applications then “contained a separate footnote on each sub-source with a brief description of his/her position or access to the information he/she was reporting.”
But as the IG’s report on CHS revealed, the FBI had no basis to track extraterritorial second-hand sources, so the CHS system “will lack complete and accurate information on its CHS coverage stemming from extraterritorial sub-sources.”
How then could the FBI determine if the sub-sources (sources for sources) were credible? Reliable? Biased? And we know from the IG’s report on FISA that even when the FBI knew that information, it was omitted from the FISA applications.
For instance, even though “Steele himself told members of the Crossfire Hurricane team that Person 1 [a sub-sub source] was a ‘boaster’ and an ‘egoist’ and ‘may engage in some embellishment’” and even though “the FBI had opened a counterintelligence investigation on Person 1 a few days before the FISA application was filed,” the FISA court was not informed of these facts.
What about the other sub-sources or sub-sub-sources besides Person 1? What negative information did the FBI learn about those individuals from Steele or from other CHSes? Were similar damning statements made about other second-hand or third-hand sources relied upon in the FISA applications for secret federal surveillance?
The FBI’s Delta system is worthless to answer these questions and the IG’s audit of the CHS validation process exposed that reality. But Horowitz’s later report on the IG’s investigation into the Page FISA applications failed to flag the import of those earlier conclusions. And the import is huge because, as the FISA applications acknowledged, “neither Steele nor the Primary Sub-source had direct access to the information being reported.”
So, as bad as the IG report on FISA abuse appeared, in reality it was much worse.