Regulatory
FSB final report lists 12 sound practices for responsible AI adoption
15 October 2026 · FSB
The FSB's final report enumerates 12 sound practices for financial institutions' responsible AI adoption, covering governance and full AI lifecycle risk management. Institutions are encouraged, not mandated, to apply them.
Points to Remember
- Practices numbered 1–12, split into AI governance and AI lifecycle risk management groups.
- SP5: assess materiality and risk of AI use cases at inception and thereafter.
- SP7: AI data must be fit-for-purpose — accurate, complete, consistent, reliable, secure.
- SP10: human oversight must match materiality, risk, autonomy, complexity and explainability.
- SP12: manage third-party AI risks including performance, data quality, supply chain, concentration and business continuity.
Why it matters
Knowing these 12 expected practices helps you frame AI controls and documentation that supervisors may ask about.
Back to Daily Updates