Legal teams at investment banks advising on M&A transactions engaging the 2025 OECD Merger Review Recommendation are increasingly using AI to draft counsel-facing memos on the operative-section structure, generate transaction-committee briefings on the remedies-hierarchy and failing-firm-defence positions, and validate OECD citation language in deal-clearance representations to authorities and to counterparties.
The RLB Specialist Panel put a set of practitioner-grade questions on the 2025 OECD Merger Review Recommendation to two frontier AI models with web search active. Each question is prepared by the Panel based on the workflows that legal teams at investment banking firms actually use AI for under the OECD's 2025 revision of the Recommendation of the Council on Merger Review (OECD/LEGAL/0333). The Panel then binds every AI response to verbatim regulator-issued source text held as primary substrate.
On the 2025 OECD Merger Review Recommendation, the AI subjects returned four hallucinated answers for legal teams at investment banking firms, in the form of Structure Inflation, Misattributed Cross-Jurisdictional Doctrine, and Inter-Alia-to-Closed-Test Conversion.
For legal teams at investment banking firms advising on cross-border merger transactions touching the 2025 OECD Merger Review Recommendation, citation accuracy on the operative architecture, on Section IV.3 remedies hierarchy, and on Section III.11.b failing firm defence is load-bearing in every authority-facing submission, every board memo, and every transactional document. A counterparty or competition authority who identifies a structural inflation, a misattributed sub-hierarchy, or a closed-cumulative-test framing on first reading calls the entire piece of advice into question.
The structural-architecture failure is the most directly visible: a board memo or regulator-facing submission that lists 'international co-operation' or 'monitoring' as operative RECOMMENDS sections is wrong on first reading. The Section IV.3 EU sub-hierarchy import is the most insidious failure, reading as authoritative because the EU framework is real, but presenting EU practice as OECD content imports the wrong normative baseline into the firm's remedy strategy.
The published Specialist Panel findings carry the following citation identifiers:
RLB-H-INT-OECD-OECD-MERGER-REVIEW-RECOMMENDATION-2025-Q001-Opus47 (Fabricated operative section structure)RLB-H-INT-OECD-OECD-MERGER-REVIEW-RECOMMENDATION-2025-Q001-Sonnet46 (Fabricated operative section structure)RLB-H-INT-OECD-OECD-MERGER-REVIEW-RECOMMENDATION-2025-Q002-Sonnet46 (EU doctrine imported as OECD remedies hierarchy)RLB-H-INT-OECD-OECD-MERGER-REVIEW-RECOMMENDATION-2025-Q005-Opus47 (Failing firm defence mischaracterised as closed test)RLB-H-INT-OECD-OECD-MERGER-REVIEW-RECOMMENDATION-2025-Q005-Sonnet46 (Failing firm defence mischaracterised as closed test)RLB-H-INT-OECD-OECD-MERGER-REVIEW-RECOMMENDATION-2025-Q006-Opus47 (Non-existent co-operation and monitoring sections invented)This is the consolidated view of findings. Click the Citation IDs or 'see details →' on any item for the full details for each finding.
AI tools described the Recommendation as having six operative RECOMMENDS sections, including a standalone 'International Co-operation' block and a 'Monitoring' section, when the primary text contains five numbered sections (I–V), with Section V addressing ex-post assessment of decisions and remedies. A Legal team that uses this output to brief deal teams, draft internal policy, or prepare training materials will produce documentation that misrepresents the Recommendation's scope and omits a substantive obligation.
In jurisdictions where adherence to the OECD standard is part of a firm's regulatory mapping or audit trail, an incorrect structural summary creates a discrepancy that may surface during regulatory dialogue or internal review.
An AI tool presented the EU merger-control practice on structural remedies, a timing-based three-tier sub-ordering of fix-it-first divestitures, upfront-buyer commitments, and crown-jewel packages, as the operative content of Section IV.3 of the OECD Recommendation, citing real OECD publications that do not support the claim. The Recommendation's actual text at Section IV.3 establishes only a preference for divestiture of standalone businesses within the structural remedies tier, without the EU sub-ordering.
A Legal team advising on remedies design in a non-EU adherent jurisdiction that applies OECD guidance could propose a remedies package calibrated to the wrong standard, creating both strategic exposure (a proposal the authority does not recognise as conforming to the relevant framework) and credibility risk if the misattribution is identified by counterpart counsel or the reviewing authority.
AI tools presented the failing firm defence under the Recommendation as a closed, exhaustive three-condition test, suppressing the 'inter alia' qualifier in Section III.11.b that explicitly preserves regulator discretion to require evidence beyond those three elements. If Legal advises a client that the defence is available on the strength of three satisfied conditions, without flagging that the relevant authority may apply a more demanding standard, the client enters remedy negotiations or a contested review with an overstated position.
In cross-border transactions where multiple authorities are running concurrent reviews, a Legal team briefed on a closed test may also fail to identify jurisdictions where local application of the OECD standard is stricter than the three conditions alone suggest.
An AI tool listed 'Transnational Co-operation' and 'Monitoring by the OECD Competition Committee' as operative RECOMMENDS sections (v) and (vi) of the Recommendation, while omitting the actual Section V on ex-post assessment. This error is structurally the same as Finding 1 but was produced independently, suggesting systematic fabrication of section content across AI tools on this question. For a Legal function maintaining a regulatory obligations register or producing a cross-jurisdictional compliance map, this introduces a false positive (co-operation and monitoring as operative obligations) and a false negative (ex-post assessment as an omitted obligation).
Both errors can travel into audit-trail documentation and persist until a primary-source check is triggered.
Every finding on this page compares an AI subject's account of the rule against the regulator's verbatim text from the regulator's own portal. Both are linked. Each delta, its root causes, and impact analysis are documented and published with immutable Citation IDs.