Login
By Lewis Kamb
In February 2023, The Wall Street Journal reported that the U.S. Department of Energy (DOE) had changed its assessment on how the COVID-19 pandemic started, concluding with low confidence based on “new intelligence” that the pandemic most likely stemmed from a lab leak.
The scoop ricocheted across national media, with news outlets, including The New York Times and Fox News, quickly producing follow-up reports built around the same core narrative — that the DOE had shifted course.
But newly released internal records now suggest that the agency’s purported change might not have been an analytical pivot based on new information, but rather an effort to clarify its original conclusions that somehow got lost in translation.
The documents — obtained by U.S. Right To Know through an ongoing Freedom of Information Act (FOIA) lawsuit against DOE — show that, in the months leading up to the agency’s reported change in thinking, DOE officials and national laboratory analysts grappled with how the department’s contribution to a 2021 report on spy agencies’ conclusions on the pandemic’s origin became misconstrued as it moved through the intelligence community, creating confusion for Congress, and later for the public.
The records indicate that, in August 2022, some Republican members of the U.S. House Permanent Select Committee on Intelligence (HPSCI) had reviewed a National Intelligence Council (NIC) report and an accompanying Q&A document prepared by DOE that stirred confusion over what the agency had actually concluded.
In turn, the NIC and Office of the Director of National Intelligence (ODNI) went back to the agency for help drafting a letter to demystify that confusion, which, at least in part, had been created by internal disputes among DOE scientists over how the agency’s original assessment was conveyed in intelligence reports.
The distinction mattered because the intelligence community’s first public report on COVID-19’s origin assessments overwhelmingly favored natural spillover, with DOE’s position described as “undecided.” The report’s outcome helped shape initial public perception that U.S. spy agencies had largely dismissed the lab-leak hypothesis, even as internal views were more varied and more technically contested.
DOE did not respond to questions emailed last week about the newly released records that reveal the uncertainty over the agency’s original assessment. ODNI did not respond to specific questions either, but issued a brief statement saying that Director of National Intelligence Tulsi Gabbard “remains committed to declassifying COVID-19 information.”
An ODNI official added the agency “is investigating intelligence failures surrounding the COVID-19 pandemic, including investigating the potential suppression of the lab leak hypothesis inside the Intelligence Community.”
Most records remain hidden within the latest 186-page tranche released by the DOE, which has long run intelligence operations that leverage its unique scientific and technological expertise to protect national security. Only a relatively small number of pages, mostly of internal email discussions, contain substantive, readable material.
The rest are heavily redacted, with the DOE claiming disclosure exemptions largely based on national security and protecting intelligence sources and methods. The redactions fully encompass the agency’s official assessment documents, captured in what analysts initially labeled a Technical Intelligence Note, or TIN.
But those records left visible reveal a sharp dispute within DOE among analysts based in three of its national laboratories that apparently contributed to later doubts over where the agency stood on the origin issue.
That dispute centered on a genetic characteristic of SARS-COV-2, the virus that spawned the pandemic, and an unusually blunt warning from a DOE analyst that an internal memo comparing the feature to naturally occurring examples in other viruses was “either a large mistake or disingenuous.”
Grappling with an assessment
In August 2021, after President Joe Biden ordered a 90-day review of intelligence on the pandemic’s roots, ODNI released a declassified summary report stating that most agencies leaned toward a natural origin with low confidence. One agency favored a lab-associated incident with moderate confidence, and two others — including the DOE — were described as “undecided.”
No agency assessment papers were released, no technical annexes were provided and no underlying intelligence was disclosed.
In June 2023, after Congress passed the COVID-19 Origin Act, ODNI issued an updated summary reflecting that both DOE and the FBI assessed, with low and moderate confidence, respectively, that a lab-associated incident was most likely.
Again, ODNI did not release any of the agencies’ own analyses or technical underpinnings.
Both summaries drew criticism from lawmakers and outside analysts who argued they were too general — describing confidence levels without providing the data or reasoning behind them, and offering the public little ability to evaluate the evidence or analytic disputes that drove each agency’s view.
The newly released DOE records suggest that this lack of detail also had a downstream consequence — confusion about what DOE had actually concluded in 2021, and whether the language used to describe it matched what DOE’s analysts had assessed.
According to the records, analysts at three of the DOE national laboratories — Los Alamos, Lawrence Livermore and Pacific Northwest — produced the agency’s assessment in 2021 for DOE’s Office of Intelligence and Counterintelligence. First referred to as a TIN, the report was later labeled a Biological Assessment Report, the records show.
That report also fed into the NIC’s classified assessment summary delivered to the president and other senior officials, and to the House intelligence committee.
As part of its contribution, DOE also prepared a Q&A document explaining its reasoning and helping translate technical findings into language intended for policymakers and overseers.
The records reflect disagreement among the DOE labs that helped prepare the assessment about whether portions of the agency’s internal analyses were too technical and misleading, but don’t indicate whether or how those disputes were resolved.
A final report for the NIC was ultimately submitted by lab analysts to DOE headquarters in October 2021.
What remains unclear is whether DOE’s analytic leaning in 2021 was later conveyed clearly in the NIC’s memo to Congress and, eventually, to the public.
Omission of a ‘flanking sequence’
One of the few places where the readable records reveal specific scientific reasoning involves a feature of SARS-CoV-2 that has drawn scrutiny since early in the pandemic: the furin cleavage site.
That cleavage site is a spot on the virus’s spike protein where a scissors-like enzyme called furin is able to snip to help the virus enter human cells. Some viruses have cleavage sites to help them infect hosts more efficiently.
The furin cleavage site in SARS-CoV-2 became a focus for scientists because the virus appeared to spread efficiently in humans early on — and because genomic comparisons showed that none of its closest known SARS relatives possessed the same feature, raising suspicions that it had been engineered in a lab.
A common argument supporting natural origins is that furin cleavage sites also exist in other coronaviruses. The records show how DOE’s Los Alamos analysts internally challenged a paper making that very case that was produced for the agency’s assessment by analysts at the Pacific Northwest lab.
The paper cited coronaviruses known as HKU1 and OC43 as examples of viruses with similar furin cleavage sites. But an unidentified Los Alamos analyst, whose name is redacted from the records, objected that the paper’s comparisons relied on too narrow a slice of genetic code — a snippet that, the analyst said, left out surrounding genetic context needed to show how distantly related those viruses were to SARS-CoV-2.
“There is lots of evidence for recombination in the SARS-CoV-2 genome,” the analyst wrote. “However this snippet of a sequence alignment gives no indication that HKU1 and OC43 are very distantly related to SARS-CoV-2 in comparison with RaTG13 and ZXC21/ZC45.”
Among the closest known SARS-like relatives at the time, neither the RaTG13 nor ZXC21/ZC45 viruses has a furin cleavage site.
“The omission of flanking sequence seems either a large mistake or disingenuous,” the analyst added.
The “flanking sequence” referred to is the genetic material surrounding furin cleavage sites that the Pacific Northwest lab’s work didn’t include. The analyst’s point was that by showing only the furin site — without the surrounding genetic context — the comparison could make the viruses look more similar than they really were.
The Los Alamos analyst’s critique also touched on another issue that has played a role in origins debates: codons, which are short genetic sequences identified by three-letter abbreviations for amino acids, the building blocks of protein.
Some scientists have argued that the specific codon pairing of CGG-CGG included in the furin cleavage site in SARS-CoV-2 is unusual for that family of viruses.
But the critique pointed out published research papers that have attempted to quantify how often the same series of codons appear in coronaviruses, suggesting that claims in the internal analysis about rarity were being overstated or misunderstood.
While the Los Alamos analyst’s critique didn’t explicitly contend that the virus originated from a lab, it shows DOE scientists arguing that the way the paper presented genetic evidence could unfairly tilt readers toward a natural-origin interpretation without giving them enough context to fairly judge the comparison.
Another email noted that the Pacific Northwest lab’s paper was “far too technical to be something we would give to policymakers as an intelligence assessment,” and suggested that “it would be best if the debate were held in open scientific literature/forums rather than in an intelligence assessment.”
How the DOE ultimately resolved those internal debates aren’t reflected in the records, but the issue appears to be what fueled the confusion in the NIC’s presentation of the agency’s assessment to Congress.
Clarifying ‘intended meaning’
The emails make clear that the push for clarification came in early August 2022 from Republican members of the Democratic-controlled House intelligence committee after a review of the NIC’s memo and accompanying Q&A document.
In response to a letter from committee members with a request to clarify their confusion, the NIC and ODNI sought DOE’s help in drafting language intended to clarify the agency’s position.
The committee seemed to be “taking things out of context,” according to an email that forwarded the letter, sent by an official whose name is redacted but is described in a message signature as the NIC’s National Intelligence Officer for Weapons of Mass Destruction, or WMD, and Biological Warfare — a position held by career intelligence analyst James Murphy at the time.
“I agree … that HPSCI is taking things out of context,” an unidentified official based at DOE’s headquarters responded, “but I do think that clarifying what we were saying in the Q&A analysis section implicated in the request is warranted.”
Not everyone within DOE agreed. One unidentified analyst who was asked to review the assessment language in question replied:
“I have read the document by the way, including the portion that you thought could be confusing, and it seemed clear to me.”
A few days later, after the NIC’s national intelligence officer offered a draft response letter to Congress, an unidentified DOE official worried that the wording needed to more directly “get at our perception of the main source of the confusion to reduce the potential that any of the IC’s efforts in good faith could be misconstrued as deliberately misleading.”
The official added that the NIC’s initial memo conveying DOE’s assessment had “meant to logically accompany deep dive briefings to properly get at the nuance of the issue, but we clearly traded too much clarity for brevity.”
With a mid-August 2022 deadline looming, a DOE official forwarded the latest version of a response memo to analysts at the Lawrence Livermore National Lab, noting that “the one thing the memo doesn’t say is what the intended meaning was for the language that is causing confusion.”
Limitations of ODNI reports
The internal grappling to clarify the DOE’s position underscores a persistent criticism about the ODNI’s summary reports: that their limited detail at best makes it difficult to distinguish how U.S. spy agencies arrived at their original judgments, and worse, cannot be trusted.
Without seeing each agency’s underlying assessments and data, policymakers, scientists and the public at large can’t learn how they’ve debated the origins issue or what evidence or uncertainties drove their confidence levels.
The revelations in the DOE’s latest records aren’t the only example of shortcomings in the ODNI summary reports.
Records obtained by U.S. Right To Know about the DIA’s intelligence assessment indicate that the agency had also found a lab-leak origin hypothesis “plausible” following serious scientific and intelligence analyses, but such information wasn’t included in the ODNI reports, both of which cite an unidentified agency believed to be the DIA as “undecided” on the matter.
Dissatisfaction over the ODNI reports led Congress to recently require, as part of this year’s National Defense Authorization Act, that Gabbard and Intelligence Community heads conduct reviews of, and make public, U.S. spy agencies’ intelligence on the pandemic’s origins.
Originally published by U.S. Right to Know.
Lewis Kamb is an investigative reporter at U.S. Right to Know. He was the first national FOIA reporter for NBC News and spent nearly a decade reporting for the Seattle Times, where he was part of the reporting team awarded the Pulitzer Prize for exposing the failures behind the Boeing 737 MAX crashes.
