Singapore’s DBS Bank (SGX: D05) has expanded the use of specialised agentic artificial intelligence tools to support roughly 1,500 relationship managers and credit risk managers worldwide in handling corporate credit assessments.The initiative, announced in mid-August 2026, follows a successful pilot involving 150 staff members.
The system deploys multiple specialised AI agents capable of managing more than 70 distinct tasks.
These agents draw information from annual reports, industry research, internal bank records and other sources to generate an initial, review-ready draft of a credit memo.
Credit assessments form a critical part of due diligence before banks extend financing to large and mid-sized corporate clients.
The process involves evaluating a company’s financial position, business outlook and risk factors.
Relationship managers have traditionally spent as much as 40 percent of their time gathering and analyzing this information and preparing the associated documentation.
DBS aims to cut the time required for these activities by at least 30 percent through the new tools.Staff members work interactively with the agents.
They can request additional research, refine the draft and integrate their own expertise on clients, industries and broader business conditions to produce the final memo.
This collaborative approach ensures human oversight remains central while the AI handles much of the data synthesis and initial drafting.
Han Kwee Juan, Group Head of Institutional Banking at DBS, highlighted the potential of the technology.
He noted that agentic AI allows the bank to capture and scale the knowledge of its strongest relationship and credit risk managers. This improves the consistency and quality of credit analysis across the organisation.
As a result, relationship managers can devote more attention to strategic discussions with clients and supporting their growth, while credit risk managers gain capacity to focus on portfolio strategy, risk calibration and emerging threats.
The rollout forms part of DBS’s wider strategy to integrate agentic AI into both internal operations and customer-facing services.
The bank describes its vision as becoming an “AI-enabled bank with a heart,” where technology amplifies human expertise rather than replacing it.
In the preceding month, DBS introduced agentic capabilities to its virtual assistants—DBS Joy for corporate clients and DBS digibot for individuals—reaching more than 10 million users across Singapore, Hong Kong and Taiwan.
Agentic AI differs from earlier generative tools by its ability to plan, reason and execute multi-step processes with a degree of autonomy.
In this application, the agents operate within carefully designed workflows and under human supervision.
Bankers retain full responsibility for final decisions and the incorporation of contextual judgement that pure data processing cannot provide.
By reducing the administrative burden of credit memo preparation, DBS expects its corporate banking teams to become more efficient and client-focused.
The move reflects broader industry efforts to apply advanced AI to complex, knowledge-intensive banking processes while maintaining rigorous risk management standards. As the technology matures, similar agentic systems may expand to other areas of institutional banking, further reshaping daily workflows for employees.
