A declassified March 2020 memo shows NSA Deputy Director George Barnes halted the release of a finished report on foreign targeting of U.S. elections, leaving important findings off the desks of policymakers for years and exposing leadership choices that favored institutional protection over timely intelligence sharing.
The memo, written by an analyst who worked on the assessment, records a March 2020 meeting where Barnes argued that publishing the report would invite accusations that the NSA was part of the “deep state.” Analysts had finished a long-delayed report covering foreign efforts targeting U.S. elections dating back to 2014, but leadership held it for months before the final decision to block it.
The document was declassified in August and released by a White House transparency review, and an administration official said it was included in the materials reviewed by Intelligence Community Ombudsman Barry Zulauf for his 2021 report. That report alleged politicization of election-interference intelligence, and this memo gives a contemporaneous snapshot of leadership choices inside NSA.
The scene the memo paints is straightforward and ugly: analysts did their work on time, the report sat with leadership for roughly sixteen months, and when the moment came to publish, the deputy director prioritized protecting the agency’s image. The result was a withheld product and a hole in what policymakers were told about foreign threats.
Barnes is recorded warning that the political climate, with an acting Director of National Intelligence tasked to “clean house” and an administration “suspicious of the IC and aggressive in removing anyone who stands in their way”, made release risky. He argued the timing—publishing anything tied to the 2016 election more than a year after the fact—would prompt questions about motive and invite charges of political targeting.
His proposed fix was to suppress the full assessment and issue a “watered-down” summary instead, a choice analysts warned would look selective. They argued that patching the product would still identify figures from the 2016 Democratic primary and that removing portions would appear even more politically driven than releasing the full report.
The memo also makes clear the report required pre-approval under the Gates Procedures, rules that govern intelligence products identifying members of Congress and related figures, which meant Director General Paul Nakasone’s sign-off was needed. Because Barnes resisted, the assessment never got to that approval phase and remained blocked.
At the same meeting, Barnes and staff discussed a separate draft about the Internet Research Agency’s preferences and strategy for the 2020 Democratic primary, which faced the same procedural hurdles. Barnes floated folding some 2016 material into the 2020 product or using a limited, non-serialized sharing process that would keep the material from broader distribution.
This episode fits a wider pattern critics have long warned about: analysts and leaders sheltering politically sensitive findings rather than getting them to decision-makers. Declassified FBI documents have separately shown officials suppressing Chinese election-interference tips, adding weight to concerns that institutional self-preservation often eclipses the duty to inform.
The report Barnes killed was not new intelligence; it covered foreign targeting back to 2014 and included material related to the 2016 election, with an earlier, narrower version already published. The fuller assessment had been proposed for release in December 2018, but repeated failures at higher levels left the product in limbo for roughly sixteen months.
By February 2020 the distribution list finally cleared and the report was ready, only to be shut down in March when Barnes raised reputation concerns. The analyst who authored the memo insisted the original language had enduring foreign-intelligence value and that suppressing it deprived policymakers and other analysts of material relevant to current and ongoing threats.
“DDIR perceived that ODNI, CIA, and to an extent the FBI had been tarred as hosting or being part of the ‘deep state’, but that NSA thus far had not been associated with the deep state and was trusted; he felt that publishing this report would destroy that trust. He felt that the questioning of NSA that would ensue would have ramifications on the credibility of NSA reporting overall and would result in morale problems among the broader NSA workforce. He repeatedly said that he wanted to protect our ability to be who we are at NSA: nonpartisan, objective, speaking truth to power.”
Read that again. Barnes framed his move as protecting the agency’s nonpartisan reputation even as he chose to withhold finished intelligence from the people it was meant to serve.
That approach is political theater wrapped in bureaucratic language. The analyst who documented Barnes’s remarks saw it that way, detailing the long trail of leadership delays that led to this choice and calling out the contradiction between the agency’s mission and the action taken to shield its public image.
Other declassified records reinforce the same concern. They show episodes where agencies filtered or delayed reports that could be politically awkward, and records have indicated Chinese actors compromised voter rolls in at least eighteen states—exactly the kind of foreign activity agencies should be reporting promptly.
The Intelligence Community Ombudsman included this memo in a broader probe of politicization, and the Zulauf report cataloged other instances where analysts let personal views affect what reached policymakers. One China analyst reportedly refused to send intelligence on Chinese 2020 election meddling to the White House, referring to the president as “that vulgarian in the Oval Office.” The community even split on assessments, with a majority view downplaying Chinese influence and a minority view backed by then-DNI John Ratcliffe reaching a different conclusion.
The memo is a documented, contemporaneous example of internal choices to withhold information in the name of protecting an institution. It raises questions about whether leaders are letting political risk guide what intelligence gets shared and whether elected officials receive the full picture they need.
The document leaves several questions unanswered: it does not say whether Barnes later approved any version of the report or whether the full assessment was ever released in another form. The analyst’s identity remains classified, as does the identity of the “acting ODNI” Barnes referenced, and the specific intelligence findings in the suppressed report remain out of public view.
Barnes wanted to protect NSA from being called the “deep state.” The memo he tried to keep hidden suggests the label fit better than he wanted to admit.
