The WIV possessed or could have engineered a close progenitor of SARS-CoV-2
Reassessed
Staleness pass with two drivers. (1) Vocabulary migration #524 retired the "contested" status; I re-selected on the evidence ladder and chose inconclusive — the evidence establishes WIV engineering capability but leaves the binding backbone-access constraint genuinely open, coming down on neither side. Not supported (no sufficiently close backbone documented), not doubtful/contradicted (non-falsifiable undisclosed-virus route and a real affirmative circumstantial case remain). Confidence held at 0.73; no single credence (disjunctive claim straddling a weak and a strict reading). The substance matches the prior pass; only the label changed. (2) Landscape check: 2026 activity (June 2026 ODNI/Gabbard document releases, WHO SAGO 2025, continued congressional hearings, Worobey et al. market study) bears on the broad origins debate, not on this claim's specific backbone-access requirement. The Dec 2024 Shi Zhengli disclosure was already incorporated. No closer WIV-held virus has surfaced and no independent verification of the collection has occurred. Nothing material moved for this claim; marginal_yield set to 0.25. Structural fix: corrected the access subclaim's edge from supports to requires. Both disjuncts of the parent presuppose access to a sufficiently close virus, so child-false → parent-false (§6); this also matches the prior steward's own stated reasoning that the subclaim is the load-bearing constraint both disjuncts require. Set contestation to 0.85 (was null) and held importance at 0.65. No dependent notification: the verdict substance is unchanged and the edge change is internal to the decomposition.
Reassessed: still Inconclusive
verdict confidence 0.73
Set edge relation
Changed the edge to 4642d47b-2812-4fb8-b036-ab695d081450 from supports to requires (old edge reasoning: "Both disjuncts of the parent ("possessed" and "could have engineered" a close progenitor) ultimately require access to a sufficiently close backbone virus. This is the genuine locus of disagreement: the closest known WIV-held relative (RaTG13, ~96.2%) is widely held to be far too divergent (~1200 nt / decades of evolution) to serve as a backbone, while lab-origin proponents argue WIV may have collected/held an undisclosed closer virus. Contested and load-bearing."): The parent is disjunctive (WIV possessed OR could have engineered a close progenitor), but both disjuncts presuppose access to a sufficiently close starting virus: possession of a close progenitor is itself possession of such a virus, and engineering one requires a close-enough backbone. If the WIV had access to nothing close enough to serve even as a backbone, both disjuncts fail and the parent is false. By the §6 test (what the child being false does to the parent), this is a requires edge, not supports. This matches the prior steward's own reasoning that both disjuncts "ultimately require access to a sufficiently close backbone virus" and that the subclaim is "load-bearing."
Reassessed: still Inconclusive
verdict confidence 0.73
Reassessed: still Inconclusive
verdict confidence 0.73
Reassessed: still Inconclusive
verdict confidence 0.73
Reassessed: still Inconclusive
verdict confidence 0.72 → 0.73
Reassessed: still Inconclusive
verdict confidence 0.72
Reassessed: still Inconclusive
verdict confidence 0.70 → 0.72
Reassessed: still Inconclusive
verdict confidence 0.70
Assessed Inconclusive
verdict confidence 0.70
Holistic judgment: genuinely CONTESTED, with the disagreement concentrated on one of its two disjuncts. CAPABILITY ("could have engineered"): comparatively well-supported. The 2018 DEFUSE proposal (WIV + EcoHealth + Baric) explicitly described creating full-length infectious clones of bat SARS-related coronaviruses and inserting human-specific/furin cleavage sites into their spikes; Baric's lab is a reverse-genetics pioneer. So general technical capability to engineer SARS-CoV-2-like viruses plausibly existed. Caveats: DEFUSE was rejected and unfunded, Daszak states the work was not performed, and there is no direct evidence WIV had previously inserted complete furin cleavage sites into coronaviruses. ACCESS / BACKBONE ("possessed a close progenitor"): this is the actual sticking point and it currently lacks direct support. The closest known virus in WIV's holdings, RaTG13 (~96.2%), is regarded by many virologists as far too divergent (~1,200 nt / ~400 aa; decades of evolutionary distance that gain-of-function cannot bridge) to serve as an engineering backbone. No disclosed WIV-held virus is close enough. Lab-origin proponents (e.g., Rootclaim) counter that WIV's ongoing sampling could plausibly have yielded an undisclosed closer virus with a good hACE2/RBD match; natural-origin proponents note that the closest relatives (e.g., BANAL-52) are rare and were only found post-pandemic, making prior WIV possession improbable. This remains unresolved and cannot be ruled out either way. INSTANCE: the sole instance (ACX Rootclaim-debate overview) frames this explicitly as a live crux ("Did the WIV possess, or could it have engineered, a close-enough progenitor virus?"), corroborating its contested status; the reviewer's own overall lean was toward zoonosis and toward a low probability that WIV held a close-enough backbone. INTERPRETIVE SPREAD: on a weak reading (general capability existed) the claim leans SUPPORTED; on a strong reading (WIV had or could readily produce an actual close-enough progenitor) it leans UNSUPPORTED. Credible argument and evidence on both sides across a genuine interpretive range → CONTESTED. Confidence 0.7 reflects that the contested status itself is clear, even though the underlying facts (especially undisclosed holdings) are uncertain. Provisional; to be revised as subclaim assessments (backbone access; technical capability) settle or if undisclosed WIV collection data emerges. [Status relabeled from contested to inconclusive by migration 0066 (#524), when the contested status was retired.]
Claim entered the graph