MAS rolls out AI risk toolkit for financial firms

The framework was built with 24 banks insurers and industry partners to guide safe deployment.

The Monetary Authority of Singapore (MAS) has announced the completion of the second phase of Project MindForge, resulting in a new AI Risk Management Toolkit for the financial sector. 

This toolkit was developed by a group of 24 banks, insurers, and other industry partners to help firms manage the risks associated with traditional AI, generative AI, and advanced "agentic" AI.

A central part of the release is the AI Risk Management Operationalisation Handbook, which gives practical steps for companies to build their own safety rules. 

The handbook is split into four main areas, starting with oversight to ensure clear roles and responsibilities for AI supervision. 

It also covers risk management to help firms identify where AI is being used and how much risk it carries, as well as lifecycle management to put controls in place at every stage of an AI’s use. 

Finally, it addresses the need for support, focusing on building the infrastructure and staff skills required to use AI responsibly.

To help companies learn from each other, the toolkit includes a collection of real-world case studies documenting the challenges and successes other financial firms have experienced.

MAS also plans to set up a new workgroup under the BuildFin.ai initiative to keep these resources updated as technology changes and to share knowledge on managing the risks of the latest AI developments.

Kenneth Gay, Chief FinTech Officer at MAS, stated that the release of this toolkit is a major step toward ensuring that AI is used safely and responsibly across the financial industry. 

The handbook will be updated periodically to reflect new technology and the expectations of regulators.