Between February and June 2026, five documented events confirmed the central analytical claims of the longevity asymmetry corpus (Huynh 2026a–n) through institutions and methodologies that share no common assumptions, no common data sources, and no common analytical framework. Chatham House’s March 2026 governance assessment concluded that coordinated preventive AI governance is “unlikely” and that effective architecture will emerge in response to crisis rather than in advance of it — the same conclusion the corpus reached through deductive structural analysis and five months earlier. Anthropic enacted a complete FM2 (Cognitive Load Externalization) cycle across two events one hundred days apart: on February 24, 2026, Chief Science Officer Jared Kaplan withdrew the Responsible Scaling Policy commitment to TIME — “We didn’t really feel, with the rapid advance of AI, that it made sense for us to make unilateral commitments … if competitors are blazing ahead” — with FM4 (Power Overrides Sovereignty) converging simultaneously as the Pentagon threatened to cancel Anthropic’s 200 million contract within the same 72-hour window; then on June 4, 2026, Anthropic published “When AI Builds Itself” proposing a coordinated global pause with a conditionality structure identical to the commitment just withdrawn, one week after filing confidentially for an IPO. The Project Glasswing case documents the governance gap in two stages: Phase 1 (April 7 to May 25, 2026), in which the European Central Bank — after six weeks of active negotiation — remained excluded from access to the vulnerability corpus Mythos had generated, confirming that regulatory authority does not confer access under private capability concentration; and Phase 2 (June 2, 2026), in which Anthropic formalized access at 25/125 per million tokens, making exclusion legible rather than merely perceptible and satisfying the visibility condition the corpus identifies as the trigger for A5 activation. The US Commerce Department’s June 12, 2026 export control directive, which required a worldwide hard shutoff of Fable 5 and Mythos 5 because targeted restriction by nationality was technically impossible, confirmed the corpus’s FM4 prediction that capability access governance would take the form of blunt unilateral state action rather than surgical multilateral coordination. This paper analyzes the epistemic structure of these five confirmations. The Chatham House conclusion is not equivalent to the corpus’s central claim: it is a lower bound. Chatham House establishes that the coordination layer of F2 governance has failed. The corpus establishes that even if coordination were achieved, the Symmetric Wound — the constitutive attachment of every governance actor to specific persons whose biological trajectory their decisions affect — makes impartial governance structurally unavailable prior to any institutional arrangement. The Chatham House conclusion confirms the floor. The corpus argues the ceiling is the floor. The Anthropic FM2 two-event sequence is not two data points but a documented cycle: existing conditional constraint removed when competitive pressure makes it costly; future conditional constraint proposed with identical conditionality structure when governance signaling creates commercial value. The ECB-Glasswing two-stage case is the only confirmation in this set that tests two structural claims rather than one: Phase 1 confirms the governance gap; Phase 2 confirms the legibility trigger — making it the most precise pre-test of the longevity access dynamics the corpus predicts, at smaller scale and with complete documentation. The most significant finding across all five confirmations is not their convergence but the second-order FM2 loop each Chatham House-class analysis illustrates: accurate institutional analysis of FM2’s prevalence, when disseminated at sufficient scale, reduces the marginal probability of the only condition — preventive governance — that would resolve it. The analysis is correct. Its dissemination makes the outcome it correctly identifies more likely. The mechanism is self-reporting.
HUYNH GIA BAO (Wed,) studied this question.