The Origins of SARS-CoV-2 · evidence ledger

Evidence

All 31 evidence records in this case's ledger. Every record carries an explicit direction, keeps what the source states separate from what we infer, and links its source and the claims it bears on. Supporting and undermining evidence get the same structure and the same seriousness.

Supports13 Undermines5 Qualifies8 Context5

Supports

13 records · strongest first

Records that favor a rung of the featured hypothesis.

Supports· strongCOV-E001

Early cases clustered around the market, including cases with no market link

what the source states: "We considered three categories of cases, which were all significantly closer to the Huanan market than expected: (i) all cases (median distance 4.28 km; P < 0.001), (ii) cases linked directly to the Huanan market (median distance 5.74 km; P < 0.001), and (iii) cases with no evidence of a direct link to the Huanan market (median distance 4.00 km; P < 0.001)." And, from the Discussion: "this group of cases resided significantly closer to the market than those who worked there, indicating that they had been exposed to the virus at or near the Huanan market. For market workers, the exposure risk was their place of work, not their residential locations, which were significantly farther afield than those cases not formally linked to the market."
what we infer: This is the observation that ascertainment bias has the hardest time explaining. If early cases were found because investigators looked around the market, the unlinked cases should not be the ones living nearest it. The comparison is significant at p = 0.029, which is a modest margin for a load-bearing result.
limitations: Rests on Chinese case ascertainment that has never been independently audited; the complete December 2019 line list has not been released.
Not independently reproduced; the code and raw geospatial data have not been re-run by a third party.
A published eLetter disputes the spatial result on distributional grounds and the authors replied identifying flaws in that critique; the exchange is unresolved.

Michael Worobey (2022)AI-verified citationCOV-C001 COV-C002

Supports· strongCOV-E004

Chinese CDC surveillance: 923 environmental samples, 74 positive, 87.5 percent of interior positives in the west zone

what the source states: "From 1 January 2020, after closure of the market, 923 samples were collected from the environment. From 18 January, 457 samples were collected from 18 species of animal … SARS-CoV-2 was detected in 74 (70 RT-qPCR and 4 Bowtie2) environmental samples, but none of the animal samples." On distribution: "Of the 64 SARS-CoV-2-positive samples collected inside the HSM, 87.5% (56/64) were collected in the West Zone of the market." On the sampling window: "The market was closed in the morning of 1 January 2020 … On the same day, in the early morning, the Chinese Center for Disease Control and Prevention … dispatched an epidemiological team." And on what was done to the site: "the outside surface of the rolling shutter doors of the stalls and the corridors were disinfected … throughout January and February 2020. The goods inside the stalls were completely cleared and disinfected until early March 2020."
what we infer: The custodian's own data independently corroborates the western-section concentration found by outside analysis, which is the strongest form of that finding. The same paper bounds what the sampling could show: the market was empty and being disinfected throughout the collection period.
limitations: Cite the 2024 version of record: the positive count was 73 in the 2023 accelerated preview and 74 after a note added in proof explaining the difference.
The paper's abstract dates animal sampling from 18 January while its Results describe both sample sets as collected from 1 January; the discrepancy is noted rather than resolved.
Produced by the custodian whose data-sharing is itself in dispute.

William J. Liu (2024)AI-verified citationCOV-C004 COV-C005

Supports· strongCOV-E005

Susceptible-wildlife DNA in the positive samples — and the authors' own limitation

what the source states: From the published abstract: investigators "find increased SARS-CoV-2 positivity near and within a wildlife stall. We identify wildlife DNA in all SARS-CoV-2-positive samples from this stall, including species such as civets, bamboo rats, and raccoon dogs, previously identified as possible intermediate hosts." From the Results: "The five SARS-CoV-2-positive samples from wildlife stall A contained mtDNA from raccoon dogs, hoary bamboo rats, dogs … and European rabbits. Amur hedgehog and Malayan porcupine mtDNA was present in four samples … and one sample contained masked palm civet mtDNA." The authors then state the limitation directly: "Although either animal or human viral shedding is consistent with these data, only a zoonotic origin of SARS-CoV-2 directly predicts co-detection of SARS-CoV-2 and wildlife genetic material." And: "It is not possible to conclude which of these species may have been infected and/or introduced the virus to the market from these data alone."
what we infer: The strongest affirmative market evidence, and its limits are stated by the people who produced it rather than by critics. The asymmetry the authors identify is genuine and easy to miss: co-detection is what zoonosis specifically predicts, even though human shedding after a zoonotic event would produce the same pattern. Detection is uneven across species, which the flat species lists in circulation obscure.
limitations: A reanalysis of the same Chinese CDC environmental data underlying the surveillance paper, sharing authors with the market epicenter paper: one evidence stream, not an independent confirmation.
The authors note that a correlational analysis of this data cannot reliably show whether any particular species was infected.

Alexander Crits-Christoph (2024)AI-verified citationCOV-C006 COV-C008

Supports· strongCOV-E010

Two lineages, two nucleotides, and the introduction-date estimates

what the source states: "Lineage A differs from lineage B by two nucleotide substitutions, C8782T and T28144C, which are also found in related coronaviruses from Rhinolophus bats." On timing: "we inferred the infection date of the lineage B primary case to be 18 November (95% HPD, 23 October to 8 December) and the infection date of the primary case of lineage A to be 25 November (95% HPD, 29 October to 14 December). The lineage B primary case predated that of lineage A in 64.6% of the posterior sample, by a median of 7 days." The corrected conclusion: "we inferred substantial support favoring separate introductions of lineages A and B (BF = 4.3 and BF = 4.2 by using the recCA and unconstrained rooting, respectively)."
what we infer: The Bayes factors quoted here are the corrected values. The originally published figures were 61.6 and 4.2 — a distinction that matters because the pre-correction numbers still circulate widely, including in material that cites this paper approvingly.
limitations: Model-based inference sensitive to rooting choices, molecular-clock assumptions, unsampled intermediates, and which early sequences are trusted.
Companion paper to the market epicenter study with overlapping authors and inputs; their agreement is not independent corroboration.
Several hundred early Chinese sequences remain unpublished.

Jonathan E. Pekar (2022)AI-verified citationCOV-C009 COV-C010

Supports· strongCOV-E014

WHO: genomic features explained by natural evolution — and not refuted as man-made either

what the source states: "evidence around the evolution of the virus suggests that these features likely developed through natural evolution in susceptible animals and humans due to a propensity for recombination and accumulation of mutations associated with coronaviruses … Furin cleavage sites occur in other human coronaviruses and had been previously identified in nature, for example in animal-associated merbecoviruses. This suggests that it could also be acquired in sarbecoviruses through natural recombination events … Based on our knowledge on coronavirus evolution and biology, this and other features of SARS-CoV-2 are readily explained by natural evolution and provide no support to theories of a man-made origin. However, the latter is also not refuted by biological data."
what we infer: The final sentence is the one that must never be dropped, and it is absent from most citation of this passage including this case's own research brief. WHO says the genomic features give no support to a man-made origin and, immediately, that biological data does not refute one. That is the precise epistemic position: the genome is not evidence for engineering, and it cannot exclude it either.
limitations: An expert synthesis rather than new analysis; it reviews published work rather than testing it.

World Health Organization (2025)AI-verified citationCOV-C011 COV-C012

Supports· strongCOV-E016

The closest known relative, at roughly 96 percent whole-genome identity

what the source states: "Simplot analysis showed that 2019-nCoV was highly similar throughout the genome to RaTG13, with an overall genome sequence identity of 96.2%." On the spike: "The receptor-binding spike protein encoded by the S gene was highly divergent from other CoVs … with less than 75% nucleotide sequence identity to all previously described SARSr-CoVs, except for a 93.1% nucleotide identity to RaTG13."
what we infer: Roughly four percent whole-genome divergence represents decades of independent evolution, which is why no sampled relative can be the immediate progenitor. Note that the paper reports identity figures and concludes probable bat origin; the "too distant to be a progenitor" inference is drawn elsewhere, and is not attributed to these authors. Later peer-reviewed papers give this same figure as 96.1 percent.
limitations: Authored by the institute whose research is itself under scrutiny in this case — a conflict to record, not a reason to discount the sequencing.
A statement about sampled diversity, not existing diversity.

Peng Zhou (2020)AI-verified citationCOV-C013

Supports· strongCOV-E021

WHO's conclusion, and its statement that the laboratory hypothesis could not be assessed

what the source states: "The weight of available evidence reviewed by SAGO suggests zoonotic spillover of SARS-CoV-2 into the human population, either directly from bats or through an intermediate host. However, SAGO cannot conclude with certainty where and when this occurred, nor if the HSM was indeed the first instance of spillover into the human population, or the site of further spillover and amplification." And: "Information and evidence is also lacking to assess the possibility of a laboratory origin - either the evidence is not available or has not been provided to the scientific community. As a result, SAGO has been unable to adequately assess this route for human infection and therefore is not in a position to rule this out as a possibility." On its own access: "It is important to note that SAGO did not have access to original raw data from any source in preparing this report."
what we infer: The two halves must always travel together. Note also the qualification "reviewed by SAGO," which ellipsis usually removes and which is the whole point: the weight of the evidence that could be seen. The report's own statement that it had no access to raw data from any source is the most important sentence in it for calibrating everything else.
limitations: The file WHO now serves is a post-publication corrected version, so page locators can differ from the June 2025 original; the version read is recorded in the source.
Not independent of the papers and assessments it reviews.

World Health Organization (2025)AI-verified citationCOV-C015 COV-C001

Supports· strongCOV-E029

The biosafety findings, with the same report's countervailing statements

what the source states: The report states that as of January 2019 institute researchers performed experiments with SARS-like coronaviruses in biosafety level 2 laboratories, and that some researchers probably did not use adequate biosafety precautions at least some of the time. The same document records that the intelligence community knows of no specific biosafety incident that caused the pandemic, characterizes a November 2019 biosafety training course at the institute as appearing routine rather than an emergency response, and describes a 2017 dissertation demonstrating reverse-genetic cloning that leaves no trace of modification.
what we infer: Opportunity is established; occurrence is not. This single document contains the strongest official support for the research-related family and the strongest official statements against it, which is why this case quotes both halves from it rather than mining one side.
limitations: A declassified summary whose underlying reporting cannot be audited.
Records drawn from this document share an independence group and must not be counted as separate confirmations.

Office of the Director of National Intelligence (2023)AI-verified citationCOV-C018 COV-C017

Supports· moderateCOV-E007

WHO's assessment of the co-location evidence, and of the testing that was not done

what the source states: "Currently, the most compelling evidence of a possible spillover from animals to humans comes from the independent metagenomic data analysis, which confirmed the presence of mitochondrial DNA of animals known to be susceptible in stalls at the HSM, and SARS-CoV-2 positive environmental swabs from the same stalls. The data confirms that these animals were present before the market was closed on 1 January 2020 and may have been a source for human infection." On the limits of the animal testing: "Despite some testing of the environment and animals sampled from the HSM, there was a limited number of susceptible species included in this testing, thus limiting the ability to make definitive conclusions. It is likely that many animals had already been removed due to public health measures imposed on 1 January 2020, when the market was shut down and disinfected; therefore, they were not traced for sampling."
what we infer: An independent expert body reaching the same two-sided reading this case reaches: the co-location is the strongest market evidence available and it establishes possibility rather than fact. The second passage is the authoritative statement of why the negative animal results are weak.
limitations: A synthesis of accessible evidence by a body that states it had no access to original raw data from any source.
Not independent of the underlying papers it reviews.

World Health Organization (2025)AI-verified citationCOV-C006 COV-C007

Supports· moderateCOV-E015

The Proximal Origin letter's two most-quoted sentences, in full

what the source states: From the introduction: "Our analyses clearly show that SARS-CoV-2 is not a laboratory construct or a purposefully manipulated virus." From the Conclusions: "Although the evidence shows that SARS-CoV-2 is not a purposefully manipulated virus, it is currently impossible to prove or disprove the other theories of its origin described here. However, since we observed all notable SARS-CoV-2 features, including the optimized RBD and polybasic cleavage site, in related coronaviruses in nature, we do not believe that any type of laboratory-based scenario is plausible."
what we infer: Quoted with their subject clauses intact, because those clauses are exactly what was later contested: "Our analyses clearly show" and "we do not believe" are claims of a different strength from the bare assertions usually excerpted. The second passage also concedes, in its own first clause, that other origin theories can be neither proved nor disproved — a concession rarely quoted alongside the conclusion.
limitations: A three-page commentary-style Letter, not a research article, despite being invoked as a study by both camps.
Its authors' contemporaneous private deliberations were later released by congressional committees; whether the published certainty exceeded private belief is a separate question this case does not grade here.

Kristian G. Andersen (2020)AI-verified citationCOV-C012

Supports· moderateCOV-E017

The funder's position: the supported viruses are too distant to be the source

what the source states: "NIH wants to set the record straight on NIH-supported research to understand naturally occurring bat coronaviruses at the Wuhan Institute of Virology, funded through a subaward from NIH grantee EcoHealth Alliance. Analysis of published genomic data and other documents from the grantee demonstrate that the naturally occurring bat coronaviruses studied under the NIH grant are genetically far distant from SARS-CoV-2 and could not possibly have caused the COVID-19 pandemic. Any claims to the contrary are demonstrably false." The same statement concedes: "To date, the origin of the SARS-CoV-2 virus that caused the COVID-19 pandemic has not been identified, despite intensive efforts to do so."
what we infer: Recorded with its interest declared: this is the funding agency's statement about research it funded. The underlying genetic-distance point does not depend on the agency's credibility — it follows from the published sequences — which is why the claim it supports is graded on the sequence data rather than on this statement.
limitations: An interested party's statement about its own conduct.
This case's research brief carried a staging-host URL for this statement; the canonical location is used here instead.

Francis S. Collins (2021)AI-verified citationCOV-C013

Supports· moderateCOV-E023

WHO's public statement on what was not provided

what the source states: The Director-General's remarks accompanying the report's release stated that hundreds of early sequences, detailed market-animal information, and Wuhan laboratory and biosafety information had not been provided, and that all hypotheses remain on the table.
what we infer: The organization's own public framing of its report, which matters because the report is frequently cited as having settled the question in favor of zoonosis. Its head said the opposite about the hypotheses' standing on the day it was published.
limitations: A statement of position, not new evidence.

World Health Organization (2025)AI-verified citationCOV-C015

Supports· moderateCOV-E027

The proposal's cleavage-site passage, at a confirmed locator

what the source states: Under the run-in heading "S2 Proteolytic Cleavage and Glycosylation Sites": "We will analyze all SARSr-CoV S gene sequences for appropriately conserved proteolytic cleavage sites in S2 and for the presence of potential furin cleavage sites. SARSr-CoV S with mismatches in proteolytic cleavage sites can be activated by exogenous trypsin or cathepsin L. Where clear mismatches occur, we will introduce appropriate human-specific cleavage sites and evaluate growth potential in Vero cells and HAE cultures." The proposal requested $14,209,245 across 3.5 project years, with performance sites including Kunming and Wuhan, China; Chapel Hill; New York; Singapore; Palo Alto; and Madison.
what we infer: A documented pre-pandemic intention, in a program with a Wuhan partner, to insert precisely the class of feature the pandemic virus carries. That is a real and uncomfortable coincidence of technique and timing, and it is evidence about what was proposed — not about what was done, which is graded separately. The non-award belongs beside the quotation wherever the quotation appears.
limitations: The text is OCR-derived from an image-only scan; two independent passes agree on every substantive word, and footnote superscripts were dropped rather than guessed.
The available copy is a third-party archive upload, not an official release or authenticated FOIA production.
The proposal was not funded, confirmed by the congressional record, contemporaneous reporting, and a negative check of federal award data.
A phrase widely attributed to this document — about a cleavage site 'able to interact with furin' — is not proposal text but a journalist's wording, and is not quoted here.

Peter Daszak (2018)AI-verified citationCOV-C016 COV-C017

Undermines

5 records · strongest first

Records that count against a rung — shown with the same structure and seriousness.

Undermines· strongCOV-E011

The 2023 erratum: support for multiple introductions cut roughly fourteenfold

what the source states: "the frequency of simulated topologies matching phylogenetic structures arising from a single introduction was incorrectly reported because of an error in the code, which has been corrected. Correspondingly, the Bayes factors in favor of multiple introductions of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) into the human population relative to a single introduction were overestimated. The text now represents the corrected Bayes factors and topology frequency (Fig. 2C), which still favors the multiple introduction scenario."
what we infer: A substantive correction to the case's most diagnostic quantitative claim, recorded as prominently as the finding it corrects. Comparing the two published versions gives the magnitude the erratum text omits: Bayes factors fell from 61.6 and 60.0 to 4.3 and 4.2, the authors' own wording moved from "strong" to "substantial" support, and the simulation frequency of the single-introduction topology rose from 0.5 to 3.1 percent. Both directions of overstatement are worth naming: the original figures should no longer be cited, and the characterization that the correction reduced the finding to negligible is contradicted by the erratum's own sentence that the corrected analysis still favors multiple introductions.
limitations: The erratum states no numbers; the magnitudes here come from comparing the two published versions of the parent paper.
Read via a dated archive snapshot because the repository deposit carries metadata only.

Science (2023)AI-verified citationCOV-C010

Undermines· strongCOV-E019

The 2023 assessment: no indication of a progenitor holding, and no engineering

what the source states: The 2023 declassified report states that the intelligence community has no indication that the institute's pre-pandemic research involved SARS-CoV-2 or a close progenitor virus, and no direct evidence of a specific research-related incident capable of causing the pandemic. It also records that almost all elements assess the virus was not genetically engineered and all assess it was not developed as a biological weapon.
what we infer: The single most load-bearing document in the case, and it is cited by both camps for different passages. Its exculpatory statements are recorded here with the same prominence as the biosafety findings it also contains, per the fairness rule. What it establishes is an absence of indication in the reporting available to the community — which is not the same as evidence of absence, and not the same as an audit.
limitations: A declassified summary whose underlying intelligence cannot be audited by any reader; the report places its own scope caveats on what it assessed.
One source doing a great deal of work in this case; records drawn from it share an independence group and must not be counted as separate confirmations.

Office of the Director of National Intelligence (2023)AI-verified citationCOV-C014 COV-C012

Undermines· moderateCOV-E003

The earliest reported case had no market link, per the first clinical series

what the source states: "By Jan 2, 2020, 41 admitted hospital patients had been identified as having laboratory-confirmed 2019-nCoV infection. … 27 (66%) of 41 patients had been exposed to Huanan seafood market." And: "The symptom onset date of the first patient identified was Dec 1, 2019. None of his family members developed fever or any respiratory symptoms. No epidemiological link was found between the first patient and later cases."
what we infer: The same paper carries evidence in both directions, which is why it is recorded once rather than split into a favorable half. Two-thirds of the earliest confirmed patients had direct market exposure — strong for the market's centrality — while the earliest case by onset had none, which is what an upstream or non-market index event predicts.
limitations: A single case's reported exposure history in an early clinical series is weak evidence about the outbreak's origin.
The identification of this patient as the earliest case, and the December 1 onset date, have been disputed in later reporting this case has not adjudicated.

Chaolin Huang (2020)AI-verified citationCOV-C003 COV-C001

Undermines· moderateCOV-E006

All 457 animal samples negative — and the species that were never sampled

what the source states: "The 457 animal samples included 188 individuals belonging to 18 species … The sources of the samples included unsold goods kept in refrigerators and freezers in the stalls … Samples from stray animals in the market were also collected, comprising swab samples from 10 cats, 27 samples of cat faeces, 1 dog, 1 weasel and 10 rats. All of the 457 animal samples tested negative for SARS-CoV-2 nucleic acid."
what we infer: Read carefully, this is much weaker negative evidence than it appears. The sampled animals were unsold stock, freezer contents, and strays — not the live wild mammals the zoonotic hypothesis implicates. The reanalysis of this very dataset states that no raccoon dogs or civets on sale at the market were ever sampled or tested, and no serology was taken from animals or handlers. An absence produced by not testing the relevant species carries little weight either way.
limitations: Sampling began only after the market was cleared, so the animals of interest had largely been removed.
The candidate intermediate-host species were not among those tested.

William J. Liu (2024)AI-verified citationCOV-C007 COV-C008

Undermines· moderateCOV-E030

The 'not found in nature' formulation is a press release's, not the report's

what the source states: The committee's press release lists as the first of five strongest arguments: "The virus possesses a biological characteristic that is not found in nature." The phrase does not appear anywhere in the 520-page report. The report's own formal finding reads: "SARS-CoV-2, the Virus that Causes COVID-19, Likely Emerged Because of a Laboratory or Research Related Accident," and its parallel five points are introduced as points a named outside author "explained" in a newspaper opinion piece. Its nearest furin-cleavage statement is a quotation of a journalist's claim that the site is "found in none of the other 871 known members of its viral family."
what we infer: A correction with two parts, both of which matter. First, the widely graded proposition — that a congressional report concluded the furin cleavage site is not found in nature — does not exist: the phrase is press-release language describing an unnamed "biological characteristic," and the report's own argument on the point is a quotation of an op-ed rather than a committee finding. Second, the underlying claim is contradicted regardless of who made it, since furin motifs occur across other coronavirus lineages. The defensible version is narrower: the site is absent from SARS-CoV-2's closest sampled sarbecovirus relatives.
limitations: A majority-party report and its press release are assertions by their authors; the companion minority report reaches opposite conclusions on the same record.
The report and the press release are treated here as two distinct sources, because their formulations differ materially.

U.S. House Select Subcommittee on the Coronavirus Pandemic (2024)AI-verified citationCOV-C011 COV-C016

Qualifies

8 records · strongest first

Records that narrow, complicate, or bound what another record can carry.

Qualifies· strongCOV-E013

The closest Laotian relatives bind human receptors efficiently — and carry no furin cleavage site

what the source states: "We found that the receptor-binding domains of these viruses differ from that of SARS-CoV-2 by only one or two residues at the interface with ACE2, bind more efficiently to the hACE2 protein than that of the SARS-CoV-2 strain isolated in Wuhan from early human cases, and mediate hACE2-dependent entry and replication in human cells … None of these bat viruses contains a furin cleavage site in the spike protein." On genome identity: "RaTG13 and BANAL-52 bat coronaviruses exhibit high nucleotide identity with SARS-CoV-2 throughout the length of the genome (96.8% for BANAL-52 and 96.1% for RaTG13)."
what we infer: A single paper carrying evidence in both directions, which is why it is recorded whole. It demonstrates that efficient human-receptor binding evolves in wild bat sarbecoviruses with no laboratory involved, which weakens one engineering argument — and it confirms that none of these natural relatives has the cleavage site, which is the observation the other engineering argument rests on.
limitations: A survey of one geographic region; absence of a feature among sampled viruses is weak evidence about its natural distribution.

Sarah Temmam (2022)AI-verified citationCOV-C011 COV-C013

Qualifies· moderateCOV-E008

WHO's scepticism about China's reported animal-testing results

what the source states: "It has been reported to SAGO and WHO that no stored samples from animals in China taken before the first identified human cases of COVID-19 (January 2020) were SARS-CoV-2-positive. This has not been independently verified by SAGO. However, it seems unlikely that there have been no animal SARS-CoV-2 reverse zoonotic infections in China since the onset of the COVID-19 pandemic (aside from Syrian hamsters in Hong Kong Special Administrative Region), considering the many susceptible species present there, the reverse zoonoses events reported in various susceptible species around the world, and the possibility of exposure to potentially infected animals at source farms and the HSM."
what we infer: Quoted at length because it is widely misreported, including in this case's own research brief, as WHO calling the missing intermediate host "surprising." It says something different and narrower: that the reported absence of ANY human-to-animal infections in China since the pandemic began is implausible. That is scepticism about the completeness of China's post-pandemic animal testing, which cuts against the reliability of the same negative results the zoonotic case is often criticized with.
limitations: An expert body's expression of doubt about another party's reported results, not an independent test of them.

World Health Organization (2025)AI-verified citationCOV-C007

Qualifies· moderateCOV-E012

WHO reports the two-introduction finding without drawing the lab-leak contrast

what the source states: "Based on phylogeny and empirical mutation frequencies, initial and updated analyses considering all currently available sequence data, Pekar et al. proposed that a single introduction of SARS-CoV-2 into the human population is 'highly unlikely'. These findings suggest two slightly different but distinct genetic virus variants: lineage A and lineage B." And, separately: "Analysis of available genomes provides evidence of two separate introductions of SARS-CoV-2 virus lineages … These findings and evolutionary analyses suggest that these introductions occurred after November 2019."
what we infer: Recorded to correct a misattribution this case had to catch. The WHO report is widely cited — including in this case's research brief — as contrasting the two-lineage pattern with the single introduction a laboratory leak would produce. It contains no such sentence. It reports the finding and attributes the "highly unlikely" judgment to the original authors. The inference that two introductions disfavor a single laboratory index infection is a reasonable one, but it is this case's inference, not WHO's, and it is recorded here as such.
limitations: A secondary characterization of the underlying analysis, not an independent evaluation of it.

World Health Organization (2025)AI-verified citationCOV-C010

Qualifies· moderateCOV-E022

Three governments, not one, withheld requested information

what the source states: "SAGO and WHO have requested further information from Member States, - including the Governments of China, Germany and the United States of America. … However, at the time of writing, all of the required information had not been provided to WHO." The three categories named are more than 500 early sequences; detailed information on the sources, locations, sampling and testing of animals sold at the market and other Wuhan markets including upstream farming or illegal trade; and information on research and field activities in Wuhan laboratories, including staff health records and biosafety information.
what we infer: Recorded because the standard framing — that China alone withheld — appears in this case's own research brief and is incomplete. Germany and the United States were also asked and did not fully comply. The three named categories are effectively WHO's own research agenda, and they map directly onto this case's decisive tests.
limitations: A statement of what was requested and not received; it does not establish why, or what the withheld material contains.

World Health Organization (2025)AI-verified citationCOV-C015

Qualifies· moderateCOV-E028

The draft margin comment, quoted whole rather than in its usual truncated form

what the source states: A margin comment appearing five times across successive drafts, attributed within the release to the proposal's lead author: "Ralph, Zhengli. If we win this contract, I do not propose that all of this work will necessarily be conducted by Ralph, but I do want to stress the US side of this proposal so that DARPA are comfortable with our team. Once we get the funds, we can then allocate who does what exact work, and I believe that a lot of these assays can be done in Wuhan as well…"
what we infer: Recorded in full deliberately. The comment is usually quoted from "Once we get the funds" onward, which removes the sentence stating its purpose — presenting a US-weighted team so the funder would be comfortable. Read whole, it evidences an intention to allocate assay work to Wuhan after an award, and an awareness that saying so up front would be unhelpful. Read truncated, it reads as concealment of a settled plan. The document supports the first reading and not the second, and the difference is one sentence.
limitations: Margin comments identify authors by Word initials only; the mapping to named individuals rests on email headers within the same release.
Drafts of a proposal that was never funded — they evidence what was drafted and discussed, not what was done.
The compilation is republished by an advocacy organization, though the underlying records are an authentic federal FOIA production.

U.S. Geological Survey / Department of the Interior (production); republished by U.S. Right to Know (2023)AI-verified citationCOV-C016 COV-C017

Qualifies· moderateCOV-E031

The minority report: both hypotheses remain plausible

what the source states: The minority members' final report concludes that a zoonotic origin and a laboratory accident are both plausible, as is a hybrid scenario, and that the viruses funded through the relevant grant are too distantly related to SARS-CoV-2 to be its progenitor.
what we infer: Recorded alongside the majority report because the two reached opposite conclusions on the same evidentiary record within days of each other. That divergence is itself informative: where a shared record supports opposite confident conclusions, the record is not determinative, and this case grades it accordingly rather than adopting either.
limitations: Assertions by their authors, not findings of fact.

U.S. House Select Subcommittee on the Coronavirus Pandemic (minority) (2024)AI-verified citationCOV-C014 COV-C015

Qualifies· weakCOV-E002

The 2024 erratum: data files corrected, results unchanged

what the source states: The authors identified duplicated data files and a pseudoreplicate count of 10,000 "rather than the n = 1000 used for all related analyses," regenerated the files and re-ran every affected test: "All results remained the same as previously reported: p < 0.001 and p-Adjusted (BH) = 0.003." The notice also records that a correction was posted to the supplementary material on May 8, 2023, and that "Science should have indexed it as a formal Erratum," which occurred on March 15, 2024.
what we infer: Recorded because corrections must be as visible as findings, and because this one is routinely conflated with the separate and substantive correction to the companion lineage paper. This erratum changed no result. Its only substantive content is a ten-month indexing delay, which is a small process-integrity observation about the publication record and not evidence about origins.
limitations: Read via a dated archive snapshot because the publisher bot-walls automated clients and the repository deposit carries metadata only.

Science (2024)AI-verified citationCOV-C002

Qualifies· weakCOV-E020

A later agency change of position, reported as resting on no new intelligence

what the source states: A spokesperson statement in January 2025: "CIA assesses with low confidence that a research-related origin of the COVID-19 pandemic is more likely than a natural origin based on the available body of reporting. CIA continues to assess that both research-related and natural origin scenarios of the COVID-19 pandemic remain plausible." Reporting at the time stated the assessment was not based on new intelligence but on a review of existing information.
what we infer: A change of position at low confidence, on a review of material already held, with no published assessment behind it. It is a real datum about where an agency stands and close to no evidence about the origin. The statement's second sentence — that both scenarios remain plausible — is routinely dropped when the first is quoted.
limitations: No agency publication exists: there is no report, no stated evidentiary basis, and no confidence rationale beyond these two sentences.
Every account traces to the same spokesperson statement — outlet duplication is one source, not many.

Central Intelligence Agency (spokesperson statement, as reported) (2025)AI-verified citationCOV-C014

Context

5 records · strongest first

Background that informs the case without pushing either way.

Context· strongCOV-E018

The intelligence split, with its confidence language intact

what the source states: The 2021 declassified assessment records that four intelligence elements and the National Intelligence Council assessed with low confidence that the first infection was most likely caused by natural exposure to an infected animal, that one element assessed with moderate confidence that it was most likely the result of a laboratory incident, and that others could not reach a judgment — together with the judgments that the virus was not genetically engineered and not developed as a biological weapon.
what we infer: The confidence language is the substance. Every position here is low confidence except one at moderate, which is the community's own statement that its evidence is deficient, inconclusive, or contradictory. Public claims that intelligence has confirmed either origin misstate this record in both directions.
limitations: A declassified summary of classified reporting that no reader can audit.
Same independence group as the 2023 report — successive products of one community, not independent assessments.

Office of the Director of National Intelligence (2021)AI-verified citationCOV-C014 COV-C012

Context· moderateCOV-E024

The 2022 preliminary report reopened the laboratory question, over a recorded dissent

what the source states: The preliminary report states that laboratory-related hypotheses remain appropriate to investigate and that further information is needed; a footnote records that three members dissented from evaluating the laboratory hypothesis.
what we infer: The document that formally reopened a question the 2021 joint study had rated extremely unlikely, which is the clearest demonstration in the case that these expert judgments move. The recorded dissent is included rather than smoothed over: disagreement within the body is part of what its conclusions are worth.
limitations: A preliminary document superseded by the 2025 assessment.

World Health Organization (2022)AI-verified citationCOV-C015

Context· weakCOV-E009

The 2021 joint study's pathway ratings, and China's unaudited animal-testing account

what the source states: The joint study rated an intermediate-host pathway "likely to very likely" and a laboratory incident "extremely unlikely," and recorded China's account of extensive animal testing with no positive results.
what we infer: Included for provenance rather than for its ratings, which later WHO reviews superseded: the 2022 preliminary report reopened the laboratory hypothesis and the 2025 assessment states it could not be evaluated at all. Citing the 2021 rating as WHO's current position is an error both of this case's research briefs specifically warn against, and it remains common in public argument.
limitations: Conducted jointly with the government whose data access was in question, on terms the study itself records.
Its animal-testing figures are the custodian's own account and have never been independently audited.
Superseded on the laboratory question by the 2022 and 2025 reviews.

World Health Organization (2021)AI-verified citationCOV-C007 COV-C015

Context· weakCOV-E025

The custodian's position: origins tracing declared complete

what the source states: The April 2025 white paper states that a laboratory leak is extremely unlikely, directs attention toward the United States, and declares China's origins investigation complete.
what we infer: Recorded as a position statement, never as evidence. Its significance for this case is structural rather than evidential: the government holding the early sequences, the market-animal records, and the laboratory documentation has declared the matter closed, which is the binding constraint on nearly every decisive test in the research agenda.
limitations: A government position paper from an interested party; it presents no new primary data.

State Council Information Office of the People's Republic of China (2025)AI-verified citationCOV-C015 COV-C008

Context· weakCOV-E026

Grant oversight failures, with no bearing on the origin question

what the source states: The inspector-general audit found that the funding agency and its grantee did not effectively monitor the relevant awards and subawards, identifying compliance and reporting failures including a materially late progress report.
what we infer: Included so the institutional-integrity layer is visible and explicitly walled off. These are real, primary-sourced findings about grant administration with zero diagnosticity for what infected the first humans. Recording them here, attached to the assessments theme rather than to any origin claim, is how this case keeps layer-three misconduct from silently feeding a layer-one verdict.
limitations: An audit of grant administration; it makes no finding about the origin of the virus and none is inferred from it.

U.S. Department of Health and Human Services, Office of Inspector General (2023)AI-verified citationCOV-C014