The growing use of artificial intelligence (AI) agents to manage payments, allocate funds and make financial decisions could pose risks to financial stability, prompting regulators and financial institutions to strengthen safeguards over how these autonomous systems access and transact through financial services.

The Monetary Authority of Singapore (MAS) said consumers and businesses must retain meaningful control over the objectives, mandates and actions of their AI agents, warning that greater autonomy must be accompanied by clear accountability and appropriate controls.

MAS managing director Chia Der Jiun said the adoption of AI agents could accelerate the movement of money and allocation of assets, potentially prompting financial institutions to impose limits on agent-directed transactions to manage risks.

“As agents move into activities with financial consequences, authority, accountability and control become essential,” he said in his keynote address at the Singapore FinTech Festival 2026 on Thursday.

Chia highlighted several challenges arising from the use of personal AI agents in financial transactions, including whether financial institutions can establish that an agent has been authorized by a customer and whether it is acting within the scope of its mandate.

There are also questions over who should bear responsibility when an AI agent executes an incorrect transaction or acts beyond the instructions given to it.

He cautioned consumers against providing personal AI agents with their financial account login credentials, saying there was insufficient assurance that such access could be provided safely.

To address these concerns, MAS is working with the financial industry to develop a trusted protocol and secure channel through which AI agents can access financial services.

The initiative forms part of its Safeguards for Agentic Finance at Runtime (SAFR) program, which seeks to establish safeguards for interactions between AI agents and financial institutions.

Chia said lessons from the first phase of SAFR highlighted four principles for safe agentic finance: bounded autonomy, clear identification and authority, specific mandates, and proportionate controls.

AI agents should be given sufficient discretion to perform useful tasks but remain within clearly defined boundaries. Their identities and authority must also be established, while mandates should specify permissible actions, limits and conditions.

In the next phase of SAFR, MAS will facilitate the development of a “know your agent” protocol to enable trusted information exchange between customer agents and financial institutions.

The protocol will cover an agent’s identity and whom it represents, the authority granted to it, the mandate under which it operates, and the assurance surrounding its origin and reliability.

This would allow financial institutions to determine the terms and scope under which an agent could access their services, as well as establish secure channels for agent transactions.

MAS will also establish an Agentic Money Sandbox with the Future of Finance Institute to explore the integration of AI agents and programmable money in corporate treasury management and payment services.

For corporate treasury management, AI agents could monitor liquidity positions and upcoming payments, determine when and where funds should be transferred, and execute permitted transactions using tokenized deposits.

For payment services, banks and payment service providers could allow customer agents to execute payment instructions using digital money, including tokenized deposits and stablecoins. Participating institutions would verify agents’ identities and authority before applying appropriate risk controls.

Beyond financial transactions, MAS is also examining how the wider adoption of AI agents could reshape financial institutions, business models, markets and the broader economy.

Chia said AI agents could significantly reduce the costs and time involved in searching, comparing, negotiating and executing transactions, potentially resulting in more frequent and faster consumer and business decisions.

Consumers could use AI agents to compare products, prices and terms across multiple providers, negotiate on their behalf and monitor price movements to secure better deals.

Businesses could also automate procurement and inventory management, while accessing a wider pool of suppliers and continuously comparing prices, delivery times and contractual terms.

However, the increased speed and frequency of transactions could introduce new risks to market competition and pricing behaviour.

Chia noted that emerging studies of agent interactions in simulated environments had identified the possibility of agent collusion, even without explicit instructions to collude. Such behaviour could result in higher prices or weaker competition.

There were also concerns that more capable AI agents could secure better prices and terms than less capable ones, potentially creating unequal access to economic opportunities depending on the quality of AI models available.

He said agent-to-agent interactions could eventually give rise to dynamic markets in which prices and commercial terms are continuously determined through negotiations between autonomous systems.

Financial institutions may therefore need to strengthen their value propositions to both customers and AI agents, while considering whether transaction limits are necessary to manage risks arising from increasingly autonomous financial activities.

Chia said regulators would need to consider the implications of AI agents’ participation in financial markets and their ability to move money at greater speed.

MAS’ approach is to work with the financial industry to develop common standards, governance practices and infrastructure to support the safe adoption of the technology.

He said regulatory clarity, clearly defined governance responsibilities and industry collaboration would be important in enabling the financial sector to adopt transformative technologies while maintaining resilience.

Ultimately, MAS aims to develop an interoperable ecosystem that enables trusted interactions between AI agents, financial institutions and digital money, while ensuring that meaningful human control and accountability remain central to the development of the agentic economy.

Singapore’s A*STAR AI Institute, MAS-established Future of Finance Institute sign AI research deal