MetaComp publishes AI agent governance framework for financial services
MetaComp, a Singapore-licensed payments company, launched the StableX Know Your Agent (KYA) Framework on April 21, a governance standard for AI agents operating in regulated financial services. The framework addresses identity verification, authorization, monitoring, and accountability for agents handling payments, compliance, and wealth management.
Financial institutions are already deploying AI agents to initiate payments and execute compliance decisions. Yet fewer than one in three organizations have adequate governance controls in place, according to McKinsey's 2026 State of AI Trust survey. Singapore businesses report higher AI adoption than global peers but only 47 percent have a documented responsible AI framework.
The accountability problem
When a human employee leaves an organization, their access is revoked. When an AI agent completes a transaction, its identity and permissions do not automatically expire. An agent can persist in a system long after its mandate has ended, with no verified identity anchor and no mechanism to intervene.
MetaComp co-president Tin Pei Ling said: "AI agents are already operating in financial services - initiating payments, making compliance decisions, managing portfolios. And yet there is no agreed standard for who those agents are, what they are permitted to do, or who is accountable when they act outside their mandate."
How the framework works
KYA operates across four pillars. Every AI agent is anchored to a verified identity linked to a real-world individual or institution. Each agent operates within strictly defined permissions governing what it can access, decide, and execute.
The framework introduces continuous, real-time monitoring of agent behavior. Risk profiles are dynamically updated as agents operate. All activities are recorded, creating an audit trail for regulators and institutions.
KYA extends governance to agent-to-agent interactions, building on the Financial Action Task Force Travel Rule. This requires the exchange of verified identity and transaction information across agent-initiated activities within a single architecture.
Compliance validation
MetaComp validated the framework across real-world transaction flows. Analysis of more than 7,000 transactions shows that relying on a single screening tool can leave up to 25 percent of high-risk exposures undetected. In an agent-driven environment, these risks multiply without defined identity layers and clear authorization boundaries.
The framework is open for adoption by financial institutions, regulators, and network partners. MetaComp developed it drawing on Singapore's Infocomm Media Development Authority Model AI Governance Framework for Agentic AI and sought feedback directly from regulators.
Broader context for management
This announcement reflects broader institutional momentum around AI governance. AI for Management professionals increasingly face questions about agent oversight, accountability chains, and risk management. AI for Finance teams must establish clear boundaries around what agents can decide and execute.
MetaComp has processed over US$10 billion in payment volume across 13 stablecoins and achieved full-year net profitability in 2025. The company raised US$35 million in Pre-A funding and holds licenses from Singapore's Monetary Authority as a Major Payment Institution.
The first skill in MetaComp's AgentX ecosystem, VisionX Know Your Transaction, packages compliance screening into a single agent-callable layer. Additional skills across cross-border payments and treasury are scheduled for late Q2 2026.
Your membership also unlocks:
