Singapore – Faster loan approval and Reduce customer onboarding time it might become the standard That’s because financial services companies around the world and in Singapore are turning to artificial intelligence (AI) agents to handle more intelligent tasks.
Experts told The Straits Times that customer service tasks such as answering questions, processing insurance claims and approving loans are some of the areas where AI agents are being tested in the financial industry.
While the technology is still rapidly evolving, early applications are already showing promise in helping bankers ease their day-to-day tasks so they can focus on higher-value work, they added.
Unlike traditional generative AI tools that require continuous prompting, agent AI or AI agents can make decisions, perform complex tasks, and manage workflows almost independently.
For example, traditional AI could explain when you are eligible for a loan. Meanwhile, Agentic AI can now evaluate customers, determine eligibility, and approve loans within hours instead of days.
This can be important given that manual data entry and paper-heavy processes tend to slow down loan approvals.
But banks can use AI agents to process documents and perform an initial risk analysis of customers before handing over files to human employees, said Dev Deep Sengupta, regional vice president for South Asia at UiPath, a global software company that develops AI and agent automation software.
For example, he noted, U.S.-based Lake Michigan Credit Union used AI agents to handle data collection and exception handling, reducing loan cycle times by 10 days. An exception on file is when your loan application or file has missing, inaccurate, or outdated information that prevents it from meeting standard approval guidelines.
Another area is intelligent credit underwriting for mortgages, auto loans, and small business loans, said Dr. Paul Beaumont, partner and data scientist at McKinsey & Company’s AI arm QuantumBlack.
In this case, AI agents can automatically aggregate and analyze applicant data from different sources.
Dr. Beaumont cited Germany’s Deutsche Bank as an example of using agent AI to achieve faster loan approvals while enhancing risk assessment by incorporating alternative data sources.
Gavin Barfield, vice president and chief technology officer of solutions at Salesforce ASEAN, said loan discovery as part of the lending process can be automated by AI agents, while human loan officers can focus on advising borrowers, building trust and finalizing loan applications.
In financial services, loan discovery refers to the process of identifying, evaluating, and applying for loan products tailored to a customer’s specific financial situation, typically through an AI-powered app or online platform.
Customer service is another area where AI agents can make an impact.
Priscilla Chong, managing director of Amazon Web Services Singapore, said insurance companies, for example, are deploying agent AI in customer interactions to accelerate claims processing.
She gave the example of Volttech, an insurance tech company based in Singapore. The company uses agent AI to power advanced text-to-speech chatbots that handle customer policy questions, process routine claims, and respond to inquiries with near-instantaneous response times.
Insurance company Singlife also partnered with Salesforce in October to launch an AI agent to improve customer service efficiency by providing faster and more accurate responses to inquiries.
This includes leveraging Salesforce’s Agentforce platform to pull in information from Singlife’s product manuals, training guides, and other materials.
Customer service representatives typically have to manually search these materials to find relevant information before responding to customer inquiries.
Singlife is also considering expanding the use of Agent AI to financial advisor representatives, the company said.
Another example is Bank of Singapore, which introduced an agent AI tool in October. We create a “source of wealth” report that details the total assets of an individual or entity and their origins, and clarifies the legitimacy of the client’s assets.
This tool reduces the time it takes to create such a report from the typical 10 days to just one hour.
As a result, bank relationship managers can now spend more time engaging with customers to better understand their financial needs and consider their portfolios.
On the security front, AI agents will improve fraud detection and response.
“By monitoring transaction flow in real time, identifying anomalous patterns, and immediately freezing compromised accounts, you can significantly reduce financial losses and protect your customers,” said Dr. Beaumont.
One of the biggest impacts of AI agents in this space is their ability to clear hundreds of thousands of alerts in seconds. He noted that it takes a human analyst 30 to 90 minutes per alert.
Dr Beaumont said AI agents have also proven useful in automating know-your-customer (KYC) processes and enhancing anti-money laundering processes.
Sengupta said AI agents can handle the heavy lifting of client due diligence by automating identity verification, entity data matching, and collection of required documents.
Future applications of agent AI in finance include autonomous market analysis and trading with minimal human intervention As well as role-specific agents who act as assistants to relationship managers and bank analysts. said the experts.
“We’re seeing banks develop entirely new products that don’t yet exist in the market,” Dr Beaumont added.
As the uses of AI agents expand, they cannot lose their human touch. Sengupta pointed out that human judgment remains important in final decision-making.
“In fact, financial services institutions are following a model where AI performs the foundational work, humans validate the results, and then AI completes the workflow,” he said.
Building trust with customers also remains fundamentally a human job, especially in areas such as asset management and financial advisory.
Chong said, “Customer relationships are built on trust, empathy, and a deep understanding of their individual circumstances. These qualities cannot be replicated by AI.”
Strategic decision making under uncertainty requires human involvement.
Even if AI provides data-based recommendations, complex, high-stakes decisions will still be left to humans, who can apply nuanced judgment and ethical considerations, Dr. Beaumont said.
