The Monetary Authority of Singapore (MAS) has concluded the second phase of Project MindForge, unveiling an AI Risk Management Toolkit intended to assist financial institutions in handling risks linked with traditional AI, generative AI, and agentic AI.
This toolkit was jointly created by MAS and 24 participating banks, insurers, and other industry stakeholders. A key part of this release is the AI Risk Management Operationalisation Handbook, which supplies actionable advice for building internal governance structures within financial institutions. The handbook covers four main areas: oversight, risk management, lifecycle management, and support.
Oversight
Oversight involves defining clear roles and responsibilities for supervising AI. This ensures that every member of the team understands their part in overseeing the use of AI technologies within their organization.
Risk Management
Risk management focuses on identifying appropriate use cases for AI and assessing associated risks, enabling financial institutions to proactively manage potential issues before they arise.
Lifecycle Management
This area addresses the controls needed during each stage of an AI system’s deployment, from initial development through to maintenance and decommissioning, ensuring that all stages are managed responsibly.
Support
Support involves establishing the necessary infrastructure and staff capabilities required for the ethical use of AI. This includes investing in training and tools to ensure staff can effectively manage and utilize these technologies.
The toolkit also features case studies derived from real-world experiences within financial institutions, highlighting both challenges faced and successful strategies employed in managing AI risks.
MAS intends to form a new workgroup under its BuildFin.ai initiative to maintain the toolkit as technology advances and to promote continuous sharing of knowledge on emerging AI risk issues. The handbook will be revised regularly to align with evolving regulatory requirements and technological developments.
The move highlights the growing focus across the Asia Pacific region on governing AI in financial services, especially concerning agentic AI systems that can make autonomous decisions. By bringing together multiple industry players to collaborate on this toolkit, MAS aims to provide practical guidance rather than theoretical frameworks.
Commenting on these developments, Kenneth Gay, Chief FinTech Officer at MAS, noted the release signifies a crucial step toward ensuring safe and responsible use of AI across financial services.
