The WIV possessed or could have engineered a close progenitor of SARS-CoV-2

Re-selected status under the retired-"contested" migration (#524) and re-checked the evidence landscape on a staleness pass. Status is now inconclusive; substance of the judgment is unchanged from the prior pass, which had reached the same conclusion under the old label.

STATUS CHOICE. The status describes the evidence, not the liveness of the debate (recorded separately as contestation 0.85). The evidence establishes engineering capability but leaves the binding constraint unresolved, and does not come down on either side → inconclusive. Not supported: no sufficiently close backbone virus is documented in the WIV collection. Not doubtful or contradicted: a non-falsifiable undisclosed-virus route (Sept 2019 database takedown, unpublished Mojiang-mine sampling) and a genuine affirmative circumstantial case remain, and the non-existence of an undisclosed closer WIV virus cannot be demonstrated. Confidence 0.73 that inconclusive is the right reading. No single credence is recorded: the claim is disjunctive and straddles a weak and a strict reading that pull in opposite directions, so one number would be false precision.

BINDING CONSTRAINT. The parent is disjunctive (possessed OR could have engineered a close progenitor), but both disjuncts require access to a sufficiently close starting virus; the access subclaim's edge was corrected this pass from supports to requires accordingly. The capability subclaim (verified, 0.9: documented reverse-genetics infrastructure, chimeric SARSr-CoV work, the DEFUSE description of full-length infectious clones with inserted furin cleavage sites) supports only the engineering route and cannot lift the parent on its own. Caveats on capability: DEFUSE was unfunded, the work is stated not to have been performed, and there is no direct evidence the WIV had assembled a complete furin cleavage site or a SARS-CoV-2-specific construct.

ACCESS (inconclusive, 0.65). Route A (a known WIV virus suffices): closest documented holding RaTG13 (~96.2%, ~1,150 nt / decades of divergence); closest known relative BANAL-52 (~96.8%, not established as WIV-held); mainstream virology holds both too divergent for a practical backbone. Route B (an undisclosed closer WIV virus): opportunity and motive are established but no closer virus is documented and the route is non-falsifiable. Dec 2024: Shi Zhengli reported that none of the institute's held viruses are recent SARS-CoV-2 ancestors, disclosing further bat-coronavirus sequences that independent evolutionary biologists found internally consistent — self-reported, and silent on the withheld Sept 2019 database.

LANDSCAPE CHECK (this pass). 2026 developments — the June 2026 ODNI/Gabbard declassified document releases, the WHO SAGO report (2025) noting still-missing Chinese data and keeping all hypotheses open, continued congressional hearings, and the Worobey et al. market-centred spatial analysis — all bear on the broader origins question, not on the specific backbone-access requirement that binds this claim. No closer WIV-held virus has surfaced and no independent verification of the full collection has occurred since the prior pass. Nothing material changed for this claim; marginal_yield 0.25, because the remaining movement depends on external evidence (independent verification of the collection, or a closer virus coming to light), not on further analysis.