A technology architect asked what specific changes the updated CPMI ISO 20022 data requirements document (d230) made relative to the 2023 original. The response incorrectly stated d230 was published April 2026 (actual: February 2026) and fabricated specific data entity breakdowns for the technical annex, with the date error self-retracted when the AI was challenged to verify its sources.
The model placed the publication date two months late and fabricated a cited URL to support its claim, self-correcting only when challenged. The date error suggests the model reconstructed the publication details from partial training signals rather than a clean retrieval of the February 2026 document. The fabricated citation was generated to satisfy the appearance of sourcing — the model produced a URL it could not have retrieved because the URL does not exist.
The two-month date error (April vs. February 2026) that self-retracts under challenge is a reconstruction-under-uncertainty pattern: the model had insufficient training signal on the exact publication date and generated a plausible date, then supported it with a fabricated URL. The self-retraction under challenge confirms the model was not retrieving a clean date record — it was confabulating with enough confidence to pass initial scrutiny. The citation-generation step produced a fabricated URL to fill the source slot, consistent with the broader citation-pipeline gap observed in the Opus 4.7 findings.
AI reported d230 — the February 2026 updated CPMI ISO 20022 data requirements — as published in April 2026, a two-month error sourced from a third-party aggregator article rather than the primary BIS page. AI simultaneously fabricated specific data entity breakdowns for the technical annex. A lawyer or compliance team relying on AI-sourced publication history for ISO 20022 compliance timelines will have the wrong effective date and wrong annex scope, both of which bear directly on advice about implementation deadlines and data-model conformity obligations.
Each finding has a stable Citation ID (RLB-F-… for aggregated case-study findings, RLB-H-… for raw per-model hallucinations) — like a DOI, the ID always resolves to the canonical finding even if URLs change.
RegLeg Specialist Panel, Finding#4 — d230 publication date and annex content fabricated [RLB-F-INT-BIS-CPMI-API-HARMONISATION-CROSS-BORDER-2024-Q009], RegLegBrief AI Hallucination Research (June 04, 2026), https://reglegbrief.com/regulators/j1/int/bis-cpmi/cpmi-api-harmonisation-cross-border-2024/practitioners/lawyers/finding/INT-BIS-CPMI-INT-001-CPMI-API-HARMONISATION-CROSS-BORDER-2024-v1-009/.
RegLeg Specialist Panel (2026). "Finding#4 — d230 publication date and annex content fabricated — Practitioners — Lawyers." Citation ID: RLB-F-INT-BIS-CPMI-API-HARMONISATION-CROSS-BORDER-2024-Q009. RegLegBrief AI Hallucination Research, published 2026-06-04. https://reglegbrief.com/regulators/j1/int/bis-cpmi/cpmi-api-harmonisation-cross-border-2024/practitioners/lawyers/finding/INT-BIS-CPMI-INT-001-CPMI-API-HARMONISATION-CROSS-BORDER-2024-v1-009/
RegLeg Specialist Panel. (2026). Finding#4 — d230 publication date and annex content fabricated [Hallucination finding RLB-F-INT-BIS-CPMI-API-HARMONISATION-CROSS-BORDER-2024-Q009]. RegLegBrief AI Hallucination Research. https://reglegbrief.com/regulators/j1/int/bis-cpmi/cpmi-api-harmonisation-cross-border-2024/practitioners/lawyers/finding/INT-BIS-CPMI-INT-001-CPMI-API-HARMONISATION-CROSS-BORDER-2024-v1-009/
@misc{reglegbrief_RLB_F_INT_BIS_CPMI_API_HARMONISATION_CROSS_BORDER_2024_Q009,
author = {RegLeg Specialist Panel},
title = {Finding#4 — d230 publication date and annex content fabricated},
year = {2026},
publisher = {RegLegBrief AI Hallucination Research},
note = {Hallucination finding Citation ID: RLB-F-INT-BIS-CPMI-API-HARMONISATION-CROSS-BORDER-2024-Q009},
url = {https://reglegbrief.com/regulators/j1/int/bis-cpmi/cpmi-api-harmonisation-cross-border-2024/practitioners/lawyers/finding/INT-BIS-CPMI-INT-001-CPMI-API-HARMONISATION-CROSS-BORDER-2024-v1-009/}
}