The Financial Stability Board has published a consultation report setting out 12 sound practices for the responsible adoption of artificial intelligence across the financial sector, opening it for comment through July 22 with a final report due in October. The practices cluster into three areas: organization wide AI governance, the management and mitigation of AI risks through the stages of development and deployment, and additional measures for AI related cyber and information security.
The development worth marking is the change in regulatory posture. As recently as last autumn the FSB framed its AI work as monitoring adoption and watching for vulnerabilities. Moving to a named set of sound practices is a step from observation toward prescription, and at its June plenary the board discussed AI adoption explicitly as a potential financial stability vulnerability. For compliance and risk leaders, that signals the era of open ended experimentation with AI is narrowing toward documented governance that supervisors will expect to see.
The original insight is that a 12 point framework from a standard setter rarely stays voluntary for long. National regulators tend to translate FSB sound practices into supervisory expectations, which means the consultation is a preview of the questions examiners will ask: who owns AI governance, how are model risks managed across the lifecycle, and how is AI specific cyber exposure controlled. The same accountability gaps that turn into enforcement, as seen when compliance failures hit fintech valuations, are exactly what this framework is trying to pre empt for AI. The practical move for institutions is to map current AI deployments against the three pillars now, while the text is still a draft, rather than retrofitting governance after the final report lands and the supervisory clock starts.
Source: Financial Stability Board.
