3\. Singapore’s Financial Sector Navigates a New AI Era as Institutions Remain Responsible for Systems Developed by Outside Providers

That arrangement can speed up adoption, but it also complicates responsibility. When a financial institution relies on an outside technology provider, customers and regulators still need to know who is accountable if a system fails or produces an inappropriate outcome.

Singapore’s Monetary Authority of Singapore has emphasized that financial institutions remain responsible for managing the risks associated with third-party AI. The approach places governance obligations on the institutions that use these systems, rather than allowing responsibility to disappear into a chain of technology suppliers.

The issue matters because financial decisions can have consequences far beyond a single digital interaction. An automated system may influence how transactions are flagged, how customers receive assistance or how information is used in a lending or compliance process.

External providers can supply valuable technical expertise. Specialized AI companies may offer models, infrastructure and tools that would be expensive for an individual bank to build from scratch. Partnerships can also help institutions experiment with new applications more quickly.

But outsourcing technology does not remove the need for oversight. A bank must understand the purpose of the system, the data it uses, its limitations and the circumstances in which its output should be reviewed by a person.

This can be difficult when AI models are complex or when several vendors contribute to a single service. Financial institutions need enough information to assess reliability, monitor performance and identify potential weaknesses, even if they do not own the underlying technology.

Data protection is another concern. Financial organizations hold sensitive information about customers, payments and business activity. They must consider how that information is accessed, stored and processed when an external AI provider is involved.

Institutions also need plans for failures. If a vendor's system becomes unavailable, behaves unexpectedly or produces unreliable results, the bank should be able to respond without losing control of essential services.

These responsibilities may influence how financial companies select their technology partners. Contract terms, security standards, testing procedures and the ability to audit performance can become as important as a product's technical capabilities.

For customers, stronger governance may be most noticeable when services become more reliable and institutions can explain how important decisions are handled. Yet responsible deployment also requires restraint: not every process needs to be automated simply because the technology is available.

Singapore's approach highlights a principle likely to become increasingly important as AI spreads across finance. Technology providers may build the tools, but financial institutions must remain answerable for how those tools are used within their services.

AI Image Disclaimer

This AI-generated illustration is for editorial visualization only and does not depict an actual bank, regulator or AI compliance system.

Sources

CNA Monetary Authority of Singapore The Business Times The Straits Times