A framework that does not specify how it could be wrong is not a framework. It is a marketing claim. This page lists the dated signals over the next eighteen months that will tell you whether the analysis on this site is right about timing, right about mechanism, or wrong about both. Read each signal as a falsification candidate. If the signal resolves one way, the framework strengthens. If it resolves the other way, the framework weakens. The page is built to be updated as resolutions arrive.Documentation Index
Fetch the complete documentation index at: https://decision-grade.ai/llms.txt
Use this file to discover all available pages before exploring further.
Three minute read. Five categories of signal. The most important is the buyer-side test: do high-stakes purchasers start demanding proof artifacts in 2026 RFPs. The substrate-level signals from regulatory cliffs in late 2026 and 2027 tell you whether the broader shift this framework sits within is propagating on schedule. Capital-market signals are noisier but informative.
The primary falsification test
This is the single signal that most directly tests the framework. If it resolves the wrong way, the framework’s timing is wrong and you should adjust accordingly. Watch: Top-20 consulting clients, the largest institutional investors, and government procurement offices. Look at their 2026 RFP acceptance criteria for analytical content.- Framework holds if even one of them requires structural-transparency artifacts (assumption registers, claim-source maps, boundary conditions, alternative explanations, or any equivalent proof artifact) as a condition of contract by the end of 2026.
- Framework weakens if all of them renew their major analytical-content contracts through 2026 without asking for proof artifacts in their acceptance criteria.
- Watch through: December 31, 2026.
Regulatory signals
Already in force as of May 2026, but worth tracking for implementation severity, waiver applications, and enforcement actions that signal whether the regulatory perimeter is real or theatrical. SR 26-2: Revised Guidance on Model Risk Management (Federal Reserve, OCC, FDIC, issued April 17, 2026). Supersedes SR 11-7. Reinforces artifact-based model validation. Watch for: enforcement actions against banks whose validation programs do not meet the revised standard. The first material enforcement action against a major bank for inadequate AI/model validation is a leading indicator that the buyer-side correction is real and propagating. GENIUS Act implementation (signed July 18, 2025). Federal framework for dollar-stablecoin reserves, attestations, and licensing. Watch for: the first OCC enforcement against a stablecoin issuer for inadequate attestation. Also watch the BDO quarterly attestation cadence for Tether USA₮ (announced January 27, 2026) for substantive vs. perfunctory content. SOC 2 2026 criteria (AICPA). Emphasize continuous risk assessment and earlier security artifacts in procurement. Watch for: enterprise procurement teams pushing the 2026 criteria into their AI-vendor RFPs. The crossover from SOC 2 (security) to AI verification is the move that brings analytical-content vendors into the artifact-based procurement perimeter. EU AI Act high-risk provisions (Regulation EU 2024/1689, enforcement timeline staggered through 2026-2027). Watch for: the first regulator-level fine against an AI verification provider for inadequate transparency or oversight. European enforcement typically leads U.S. enforcement by 12-18 months in adjacent domains.Substrate signals
These signals come from outside the analytical-content market but bear directly on whether the broader verification-collapse regime described in Perera and Wickramasinghe’s The Verification Collapse is propagating on schedule. If their substrate-level predictions resolve as expected, the firm-level timeline this framework specifies likely accelerates. November 10, 2026: MOFCOM rare-earth suspension expiry. China’s Ministry of Commerce holds a one-year option to reinstate the October 2025 export-control package on rare earths, lithium batteries, graphite anodes, and related processing technologies. Framework holds if Beijing reinstates (the integration race accelerates and procurement signals propagate faster). Framework weakens if Beijing extends without conditions and Western diversification stalls. May 2027: AUKUS Pillar Two operational deliverables. The first publicly-visible test of whether the trilateral defense-technology partnership produces fielded advanced-capability deliverables on the specified timeline. Framework holds if at least one operationally visible deliverable enters service. Framework weakens if Pillar Two yields zero deliverables fielded by the gate. June 30, 2027: Section 1260H indirect-procurement cutover. FY24 NDAA Section 805. The legally complex of the three procurement cliffs, with a component exception. Watch for: whether the DoD enforces the cutover meaningfully or grants broad waivers. Wholesale Chinese Military Companies List delistings before this date weaken the framework. Visible enforcement strengthens it. October 1, 2027: CATL battery restriction. FY24 NDAA Section 154. Named-entity prohibition on DoD procurement from CATL, BYD, Envision, EVE Energy, Gotion, and Hithium. The sharper of the two component-specific cliffs. Watch for: waiver applications and the rate at which Western battery-substitution capacity comes online. December 23, 2027: Section 5949 semiconductor cutover. FY23 NDAA Section 5949. Federal executive-agency procurement prohibition on covered semiconductors traceable to SMIC, CXMT, YMTC, or affiliates. Watch for: the Federal Acquisition Regulation Council’s final rule and whether it preserves the cutover or softens it with broad waivers.Capital market signals
Noisier than the regulatory signals but informative about how capital markets are pricing the substrate transition. Anduril Series H pricing. Reported in marketing since February 2026 at targeted post-money above 45 billion post-money or fails to close. Helsing trajectory. Bundeswehr framework contract was compressed from €4.3 billion to €2 billion in February 2026 following Helsing-STARK Virtus testing outcomes. Watch for: subsequent Bundeswehr trial outcomes and whether European autonomy-stack capital formation continues at the current ~25% of American levels or compresses further. Public-market autonomy names. AeroVironment (down ~57% from 52-week high as of early 2026), Kratos (down ~55%), DroneShield (recovering through Q1 2026 with A$74.1M revenue). Watch for: whether the divergence between private autonomy-stack valuations and public legacy-adjacent prices persists, narrows, or inverts. Stablecoin Treasury demand. Tether’s exposure to U.S. Treasuries exceeded 2 trillion stablecoin market capitalization by 2028. Material deceleration weakens the monetary-substrate component of the broader thesis.Technology and capability signals
Less time-bound than the regulatory cliffs but worth tracking for inflection points. Late 2026 or early 2027: powerful AI emergence. Anthropic’s stated expectation in their OSTP submission. Watch for: the first independently-verified general-capability inflection. The framework’s verification-deficit thesis intensifies sharply at any such inflection. Directed-energy operational scale. Iron Beam reached operational status with the IDF on December 28, 2025. The IDF target is fourteen Iron Beam batteries for national operational impact. Watch for: that threshold being reached, and the first non-U.S./non-Israel allied directed-energy battery operational capability. Deepfake-based fraud growth. January 2024 Arup engineering-firm fraud of approximately $25 million via AI-generated executives in a Hong Kong video conference. JINKUSU CAM and successor instruments are now targeting financial-services liveness checks. Watch for: the first board-level corporate liability event attributable to deepfake-based fraud at a Fortune 500 company. That event activates the identity-verification analog of the analytical-content correction.How to use this page
Review monthly
Resolved signals get marked. Pending signals stay on the watchlist. New signals get added as the framework develops.
Track resolution direction
Each signal that resolves contributes either to “framework holds” or “framework weakens.” Keep a running tally. A framework that is consistently being weakened by its own falsification candidates should be revised, not defended.
Reassess timing at each gate
The MOFCOM decision in November 2026 is the first major gate. If the rare-earth suspension is extended, the broader substrate transition is delayed and the firm-level timing this framework specifies probably extends correspondingly.