IOSCO Supervisory Toolkit for AI Use in Capital Markets
Final report FR/02/2026, published 25 May 2026 by IOSCO's Fintech Task Force after two years of work. Provides supervisors with practical, non-binding, non-prescriptive tools applicable across regulatory models, covering the full lifecycle of an AI system and all system types — from traditional machine learning through generative AI to emerging agentic techniques. A standalone extract of the toolkit is published separately as OR/07/2026 for use during inspections.
Why it matters for autonomous finance
Written for the supervisor rather than the supervised, which makes it the most useful document on this list for anyone trying to anticipate examination. It sets out areas of supervisory consideration, then indicators and data sources — including on-site inspections, targeted surveys, documentation requests and supervisory dialogue. If you want to know what you will be asked for, this is the closest published approximation. It also, unlike SR 26-2, explicitly includes agentic techniques in scope.
What it does not cover
IOSCO sets no binding rules; member jurisdictions adapt it to their own legal frameworks, and adoption will be uneven. Being non-prescriptive is the point and also the limit — it tells a supervisor what to ask, not what answer is adequate.
No conformance definition, and no position on ownership of agent output. Every entry in this register carries this section. A standard read past its scope is worse than no standard, because it produces confidence without coverage.