Germany Sub-Threshold: EU Return-Hubs Deal and Court-Ruled-Unlawful Border-Rejection Policy — New RETURN-HUBS-AND-NON-REFOULEMENT Ruling

Deportations to Afghanistan and Syria are limited to convicted criminals under judicially supervised, case-by-case proceedings that Amnesty International and ECRE contest on non-refoulement grounds — a real but limited EQU universality concern. The sharpest signal is the government's continuation of a border-rejection (pushback) policy that a German court ruled incompatible with EU law — an INT and ACC deficit. Critically, this conduct is contested through functioning, independent courts (the inverse of India's judicial-override pattern), so the assessed downgrade is modest: EQU 3.5→3.2, ACC 3.8→3.6, INT 3.5→3.3, formula basis 77.1→74.9 (−2.2), below the 5-point threshold and non-band-crossing (stays Established). No change proposal. A new methodology ruling — RETURN-HUBS-AND-NON-REFOULEMENT — is established, explicitly distinguishing Germany's contested-courts posture from the ACTIVE-STATE-PERPETRATION-REWEIGHT anchor (India 15.6): functioning rule-of-law check plus judicially supervised criminal-only deportations versus a directed mass-refoulement regime of UNHCR-registered refugees and citizens. A separate material data-hygiene discrepancy is confirmed: stored composite 72.8 does not match the canonical formula output 77.1 (diff 4.3, validator-confirmed warning). The assessment was computed on the formula-consistent basis; the published index was not modified.

The program is EU-exempt because GDPR prohibits it; US regulatory attention from the FTC and state attorneys general is anticipated but not yet filed. This is a genuine I3 (internal consistency) and B5 (consent orientation) labor-privacy harm. However, it reinforces an already near-floor internal-consistency baseline (INT 1.8, ACC 1.5) rather than revealing previously-unpriced harm, and is announced and deploying — not yet audited or adjudicated. The defensible adjustment is below 1 point, sub-threshold and non-band-crossing. The new CANDIDATE-INTERNAL-CONSISTENCY-SIGNAL framing documents the conduct pattern as a standing forward trigger: a formal regulatory action or an independent audit establishing scope and harm converts this to a scored INT/BND downgrade.

Tonight's evidence — BJP West Bengal election win followed immediately by a detention-centre order; continued May–June expulsions; 40 Rohingya placed on a vessel and brought to sea; Gujarat 10,000+ demolitions; USCIRF 2026 Report of Concern designation — is corroborating continuation of the same government-directed refoulement-and-expulsion regime documented in the 06-01 proposal. No new magnitude; no duplicate proposal generated. The 06-01 proposal remains the authoritative pending item at 15.6. The nationwide Special Task Force deportation directive is active; functioning courts continue to contest individual acts, sustaining the score above the absolute floor.

The remote-reclassification-as-WARN-avoidance theory and New Jersey state WARN proceedings remain filed-but-unadjudicated and are not independently scorable under the established FILED-BUT-UNADJUDICATED-LITIGATION ruling. The entity sits 0.6 points above the Critical boundary; any adjudicated or corroborated finding on the WARN-avoidance theory in NJ state proceedings or the Washington class action before or after June 15 constitutes an immediate scoring trigger from a boundary-sensitive baseline.

• The June 1 EU return-hubs deal is not yet operationalized — third-country agreements targeted by end of 2026; not scorable under announced-not-operationalized standard
• Germany's continuation of a court-ruled-unlawful border-rejection policy is the sharpest current non-refoulement signal among EU member states, assessed at −2.2 formula basis, sub-threshold
• The new RETURN-HUBS-AND-NON-REFOULEMENT ruling creates the explicit anchor distinguishing contested-courts posture (Germany) from active-state-perpetration (India 15.6)
• Austria, Netherlands, Denmark, and Greece are co-signatories to the return-hubs framework and are on sector watch for operationalization outcomes
• Germany's data-hygiene discrepancy (stored 72.8 vs canonical 77.1, diff 4.3) is a genuine validation flag requiring reconciliation before any future scored proposal

• xAI/Grok at 0.0 receives the cycle's most consequential reinforcing evidence: DoD responsible-AI chief resignation, classified-access override, below-standard safety alignment, reported CSAM-adjacent editing, SE Asia bans — all pointing the same direction
• Meta AI's MCI no-opt-out surveillance program introduces the CANDIDATE-INTERNAL-CONSISTENCY-SIGNAL framing; a formal regulatory action converts this to a scored INT/BND downgrade from near-floor INT 1.8
• Cohere's motion-to-dismiss denial (Nov 2025, substitutive-output and trademark-hallucination theories) is the leading AI-copyright reassessment trigger; MTD denial is procedural, not merits — reassess on summary judgment or liability
• OpenAI net-neutral at 27.5: DoD classified-deployment guardrails are reference-to-law-only, offset by the May 28 Frontier Governance Framework; comprehensive full assessment is overdue by 45+ days
• EU AI Act August 2 GPAI enforcement is the most proximate systemic scoring event for the full AI labs cluster — 61 days out with no published compliance documentation from any assessed lab