The Financial Stability Board (FSB) has finished its guidance on how banks, insurers and other financial firms should bring in artificial intelligence. On October 15, 2026 it published the final report, titled "Sound Practices for Responsible Adoption of Artificial Intelligence (AI)", according to a note from Japan's Financial Services Agency dated October 16.
What the report contains
The report sets out 12 sound practices meant to encourage responsible AI adoption by financial institutions. The FSA's note says they address two areas: AI governance across the whole organisation, and a comprehensive risk-management framework that covers the AI lifecycle. The FSB is the international body that coordinates financial regulators and standard-setters.
The FSA note says the practices are meant to promote responsible adoption. It does not describe them as binding rules, and it announces no Japanese rulemaking or supervisory deadline tied to them.
What is missing from the FSA note
The FSA note does not list the 12 practices or say how they divide between governance and lifecycle risk. It points readers to the FSB's publication page, its press release and the report itself, all hosted on the FSB website. Firms that want the specific expectations on offer will need the FSB report, not the FSA's summary.
The FSA filed the item under its international relations pages for the FSB, as an international standard-setting output. For Japanese financial institutions, the practical question is how the 12 practices line up with the AI controls they already run.