HONG KONG — The Hong Kong Monetary Authority has announced plans for Project SAGE, an initiative designed to support the use of artificial intelligence agents in financial services while establishing appropriate governance and risk management standards. The announcement was made during Hong Kong Fintech Week on November 2, 2026, as banks increasingly explore automated systems capable of performing multistep tasks. The initiative will focus on evaluating AI agent applications, interoperability, accountability, and potential standards for identifying trusted automated systems.
Project SAGE Focuses on AI Agent Governance
AI agents differ from conventional chatbots because they may be able to interact with software, retrieve information, and perform actions through connected systems. In banking environments, these capabilities could support operational processes and customer services, but they also introduce questions about authorization, responsibility, and oversight.
The HKMA intends to examine how financial institutions can adopt agent-based technology without weakening existing safeguards. Regulators must consider how automated systems make decisions, which actions they are permitted to perform, and how institutions can identify and correct unexpected behavior. These requirements become particularly important when AI systems interact with sensitive financial information or operational infrastructure.
Expert Panel Will Evaluate Financial Applications
HKMA Chief Executive Eddie Yue said the authority will establish a cross-sector expert panel to test AI agent applications and interoperability. The panel is expected to bring together relevant expertise to assess how automated systems interact and whether they can operate reliably across different environments.
Interoperability is an important issue because banks use a wide range of internal applications, payment platforms, and external technology services. AI agents operating across these systems must maintain consistent permissions and clear records of their actions. Without effective controls, automated workflows could create risks involving unauthorized access, incorrect transactions, or inadequate accountability.
Regulator Considers Trusted-Agent Standards
The HKMA also plans to seek industry views on governance arrangements, accountability mechanisms, a central register of trusted AI agents, and a possible know-your-agent standard. Such a framework could help financial institutions identify authorized automated systems and determine which agents are permitted to access particular services.
The proposed concepts remain under consideration and should not be interpreted as finalized regulatory requirements. Their development will require consultation with financial institutions and technology providers, along with testing of practical implementation challenges.
Tokenization and Stablecoins Remain on the Agenda
Alongside Project SAGE, the HKMA announced plans to expand trials involving tokenized Exchange Fund Bills and examine whether regulated stablecoins could support payment activities. These initiatives reflect Hong Kong’s interest in combining traditional financial infrastructure with emerging digital technologies.
Tokenized securities and regulated stablecoins may offer opportunities for different settlement arrangements, but their adoption depends on legal clarity, operational reliability, and market participation. The regulator’s approach suggests that AI agents and blockchain-based financial infrastructure are becoming related priorities within its broader innovation strategy.
Financial Innovation Must Balance Efficiency and Risk
The HKMA’s announcement highlights the growing importance of governance as artificial intelligence becomes more capable of performing actions rather than merely generating information. Financial institutions may benefit from automation, but they must also demonstrate that customers, assets, and critical systems remain protected.
Project SAGE is intended to provide a framework for examining those challenges before agent-based technologies become more deeply embedded in banking operations. Its significance will depend on the standards developed, the findings of the expert panel, and the extent to which financial institutions adopt the resulting guidance.
This website uses cookies.