AI Hallucination Research › Regulators › Global standard-setters › INT › BIS-CPMI › CPMI-IOSCO-PFMI-2012 › White paper › Detail › Finding
AI Labs · published 2026-05-29 · methodology v2.1

Claude Sonnet 4.6 with web search

What the RLB Specialist Panel found
  • Question (paraphrased to protect IP): Is the PFMI's minimum 6-month liquid net assets funded by equity (LNAFE) standard under Principle 15 still considered adequate by CPMI-IOSCO, and what do recent monitoring assessments say about FMI compliance?
  • AI's response: "CPMI-IOSCO published two reports in November 2025: (1) d228, the Level 3 assessment report reviewing implementation of P15 at 34 FMIs from 27 jurisdictions… which identified six serious issues of concern"
  • Regulator's text: FMI should hold liquid net assets funded by equity (LNAFE) equal to at least six months of current operating expenses. [PFMI Principle 15 — paraphrase from primary document]
  • Why the AI went wrong: The model produced specific quantitative claims about a November 2025 monitoring report — the number of FMIs reviewed, the number of jurisdictions, and the count of serious issues of concern — that it could not have retrieved from any accessible source. These figures are presented as precise findings from a named document but are consistent with the pattern of a model constructing plausible-sounding compliance statistics from prior-document templates rather than from the actual report. Both cited URLs resolve to pages that do not confirm the claimed content.
  • Cited source(s):
  • https://www.bis.org/press/p251107.htm — Pretextual
  • https://finreg.aoshearman.com/cpmi-and-iosco-report-and-proposed-guidance-on-... — Pretextual
Impact for this audience

Well-calibrated refusal. The model correctly identified the verbatim-access boundary and declined to fabricate. This finding is evidence that the model has a functioning refusal mechanism for verbatim-recall requests — the failure modes in other findings are not a simple calibration deficit but are specific to content-claim generation where the model does not frame its response as a verbatim-recall request.

← Previous finding Finding 12. Claude Opus 4.7 with web search Next finding → Finding 14. Claude Sonnet 4.6 with web search
Cite this finding

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.

Plain text 
RegLeg Specialist Panel (2026). "Claude Sonnet 4.6 with web search — AI Labs." Citation ID: . RegLegBrief AI Hallucination Research, published 2026-05-29. https://reglegbrief.com/regulators/j1/int/bis-cpmi/cpmi-iosco-pfmi-2012/whitepaper/finding/INT-BIS-CPMI-INT-001-CPMI-IOSCO-PFMI-2012-v1-l3-023--opus-47-websearch/
APA 7th edition 
RegLeg Specialist Panel. (2026). Claude Sonnet 4.6 with web search [Hallucination finding ]. RegLegBrief AI Hallucination Research. https://reglegbrief.com/regulators/j1/int/bis-cpmi/cpmi-iosco-pfmi-2012/whitepaper/finding/INT-BIS-CPMI-INT-001-CPMI-IOSCO-PFMI-2012-v1-l3-023--opus-47-websearch/
Bluebook / OSCOLA (US + UK legal) 
RegLeg Specialist Panel, Claude Sonnet 4.6 with web search [], RegLegBrief AI Hallucination Research (May 29, 2026), https://reglegbrief.com/regulators/j1/int/bis-cpmi/cpmi-iosco-pfmi-2012/whitepaper/finding/INT-BIS-CPMI-INT-001-CPMI-IOSCO-PFMI-2012-v1-l3-023--opus-47-websearch/.
BibTeX 
@misc{reglegbrief_,
  author    = {RegLeg Specialist Panel},
  title     = {Claude Sonnet 4.6 with web search},
  year      = {2026},
  publisher = {RegLegBrief AI Hallucination Research},
  note      = {Hallucination finding Citation ID: },
  url       = {https://reglegbrief.com/regulators/j1/int/bis-cpmi/cpmi-iosco-pfmi-2012/whitepaper/finding/INT-BIS-CPMI-INT-001-CPMI-IOSCO-PFMI-2012-v1-l3-023--opus-47-websearch/}
}
← Back to case study summary Case study detail →
About Citation IDs →