Singapore proposes governance framework for AI agents in finance

The Monetary Authority of Singapore (MAS) has published an industry white paper proposing a governance framework for AI agents operating in financial services. The framework was developed in collaboration with leading financial institutions and fintech companies.

Titled Safeguards for Agentic Finance at Runtime (SAFR), the framework aims to ensure AI agents carrying out financial tasks autonomously operate safely, securely and reliably within mandates, policies and risk limits defined by financial institutions.

SAFR addresses a central challenge of agentic AI in finance: autonomous systems increasingly operate at a speed and scale that make real-time human intervention impractical. It introduces governance checkpoints that verify and record an AI agent’s proposed actions before execution, incorporating policy-bound execution, real-time validation, auditability and interoperability directly into operational workflows.

Industry participants have already applied the framework to several use cases, including AI agents executing routine payments and treasury transactions within predefined limits, reviewing documents and generating compliance assessments for wealth management, and drafting client communications within approved content boundaries.

MAS has invited industry partners to join its BuildFin.ai working group to help shape future versions of SAFR. The recently established Future of Finance Institute will support adoption through industry pilots and regulatory sandbox experiments, with expressions of interest open to institutions wishing to test SAFR-aligned solutions.

Why does it matter?

As AI agents begin executing transactions, assessing compliance and interacting directly with customers, financial institutions need governance mechanisms that operate at machine speed rather than relying solely on human oversight. SAFR represents one of the first practical frameworks designed to embed policy checks, validation and auditability into AI-driven financial processes before actions are carried out.

The framework also reflects Singapore’s collaborative approach to financial innovation. By developing governance standards jointly with banks and fintech companies, MAS is seeking to create safeguards that are both technically practical and easier for the industry to adopt, potentially providing a model for other financial regulators exploring agentic AI.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!