Singapore financial institutions face independent review for AI use cases

Pre-deployment review by parties outside development

The Register reports that the Monetary Authority of Singapore (MAS), the country's central bank and financial-sector regulator, published its first Guidelines on Artificial Intelligence Risk Management for Financial Institutions. The guidelines reportedly ask regulated financial institutions to have parties uninvolved in development review each AI use case and its systems or models before deployment.

They also call for technology and cybersecurity reviews, ongoing monitoring, and accountability for third-party AI providers. For high-risk AI applications, institutions are reportedly expected to prepare fallback options, including alternative systems or manual processes.

The available report does not provide a comparable prior regulatory requirement for Singapore financial institutions' AI use cases.

The reported guidelines call for independent pre-deployment review of every AI use case, along with continuing monitoring, AI inventories, supplier oversight, and contingency arrangements for high-risk uses. They are due to take effect on October 7, 2027.

Ongoing monitoring and fallback arrangements

The reported guidance would apply controls not only to internally developed AI but also to third-party AI used in financial services. It also links deployment approval to continuing monitoring and contingency planning, rather than treating pre-deployment assessment as the only control.