Singapore’s largest bank is transforming its virtual assistants from information tools into digital agents that can actually get things done.
Let me show you something.
On July 28, 2026, DBS Bank rolled out an agentic artificial intelligence assistant to 350,000 corporate clients in Singapore, including small and medium-sized enterprises . Through DBS Joy, corporate clients can now retrieve and analyse account information within conversations, eliminating the need to manually search for transaction details across the platform .
This isn’t just a chatbot upgrade. It’s a fundamental shift in how banking services are delivered — moving from helping customers find information to helping them get things done .
DBS Bank is setting a new standard for AI adoption in the financial sector.
Table of Contents
- What Is Agentic AI and Why Does It Matter?
- DBS Joy: From Information Tool to Banking Agent
- The Numbers: Adoption and Impact
- What’s Next: Expansion and New Capabilities
- The Governance Challenge: Why DBS Is Holding Back
- Beyond Corporate Banking: DBS Digibot and Wealth Management
- The Bottom Line: AI Is Driving Real Business Value
- FAQ
DBS Bank’s Agentic AI: What It Is and Why It Matters
Understanding DBS Bank’s agentic AI strategy requires knowing what agentic AI actually means.
DBS Bank’s agentic AI strategy is transforming how corporate clients interact with their finances.
Agentic AI refers to systems that can understand a goal, plan a sequence of steps, and carry them out — rather than simply producing text on request . Unlike traditional chatbots that provide information or instructions, agentic AI can actually act on behalf of customers .
DBS Group Chief Operating Officer Derrick Goh explained the shift: “The true value of AI lies in delivering meaningful outcomes for customers at scale… We are taking the next leap forward – moving from helping customers find information to helping them get things done” .
DBS Chief Data and Transformation Officer Nimish Panchmatia provided a concrete example: before agentic AI, a corporate relationship manager might spend days working through a 120-page annual report, a sustainability report, and the client’s financials. Now, a chain of 70 to 80 agents pulls that material together, including external news and internal notes, before delivering it in a pre-formatted structure .
This is especially significant in corporate banking, where digital assistants remain uncommon and agentic capabilities even more so .
DBS Joy: From Information Tool to Banking Agent
DBS Bank’s DBS Joy virtual assistant is at the center of this transformation.
DBS Bank has been investing in AI for more than a decade and sees it as a strategic differentiator.
DBS Joy was launched in 2018 and revamped with generative AI capabilities in 2025 . With its new agentic capabilities, DBS Joy can now:
- Retrieve and analyse account data within a single conversation
- Answer natural-language questions like “Did my payment to ABC company go through?” or “How much did I pay in fees?”
- Execute selected banking requests without customers needing to navigate multiple screens
The agentic capabilities are activated only after customers log in to DBS IDEAL (the bank’s digital banking platform for businesses), with the AI acting solely on instructions initiated by authenticated customers .
New live chat capabilities provide direct access to a human agent when needed, and the ability to seamlessly transition from a chat to a phone conversation will be introduced in the fourth quarter of 2026 .
The Numbers: Adoption and Impact
The adoption numbers show that DBS Bank’s AI strategy is working. The adoption data also shows DBS Bank’s AI strategy is delivering measurable business value.
The data shows that customers are embracing DBS Joy:
| Metric | Change |
|---|---|
| Chats handled (H1 2026) | +22% |
| Active users | +61% |
| Calls/emails to customer service | -7% |
| Customer satisfaction | +17% |
In the first six months of 2026, DBS Joy saw a 22% increase in chats and a 61% increase in active users . This contributed to a 7% reduction in calls or emails to customer service .
Customer satisfaction scores for DBS Joy rose by 17% over the same period .
In early trials of the technology, which began in February 2025, DBS Joy handled over 120,000 unique chats from around 4,000 corporate clients each month, the vast majority of which are small and medium enterprises .
What’s Next: Expansion and New Capabilities
DBS Bank is expanding agentic AI to Hong Kong and other markets. DBS Bank expects DBS Joy to handle more than one million chats monthly across its network.
DBS isn’t stopping with Singapore. The corporate agentic AI capabilities will be extended to another 100,000 users in Hong Kong in September, and then progressively rolled out to other key markets .
The bank is also:
- Deepening DBS Joy’s capabilities and knowledge of its SME banking solutions, products, and flagship programmes to better support SMEs
- Adding voice capabilities later this year
- Introducing agentic features for the retail virtual assistant, DBS digibot, in the fourth quarter of 2026
DBS expects both DBS Joy and DBS digibot to handle more than one million chats monthly, with a total reach of 10 million customers .
The Governance Challenge: Why DBS Is Holding Back
DBS Bank is holding back on full autonomy due to governance concerns.
DBS Bank’s governance approach to agentic AI shows the technology is moving faster than safeguards.
Despite the progress, DBS is refusing to let AI agents act without human oversight. The reason: the technology for supervising agents is nowhere near as mature as the technology for building them .
Nimish Panchmatia said: “The speed at which innovation is happening in terms of capability is, for argument’s sake, say 5x. The speed of innovation of governance and control and management of these agents is at 1x. So, there’s a gap and we need to close this gap before we allow autonomy” .
The concerns are serious:
- Unexplainability: Building sprawling agents that handle dozens of tasks in sequence makes failures harder to trace
- Behavioural risk: Agents optimize for efficiency, not compliance. “The technology is designed in such a way that it will find the easiest path. The easiest path may not be the right path,” Panchmatia said
- Regulatory stringency: Banks have far more stringent requirements than other sectors like travel or retail
As such, DBS caps the number of steps any single agent chain performs, with checks applied between stages. Every deviation from expected behaviour must be flagged and the process halted for review .
Beyond Corporate Banking: DBS Digibot and Wealth Management
DBS Bank is also expanding AI across its retail and wealth management businesses.
DBS is also expanding AI across its retail and wealth businesses.
DBS digibot, the group’s retail virtual assistant, is now using generative AI to help retail customers with card-related queries, refunds, fee waivers and remittances. In the first half of 2026, the chatbot resolved nine in 10 queries digitally .
In August 2026, DBS digibot will be integrated into digiWealth, the bank’s wealth management platform, to help mass-market and emerging-affluent customers navigate investment-related questions .
More agentic AI features – such as checking card usage, tabulating card reward points, requesting fee waivers, and blocking or replacing cards – will be incorporated into DBS digibot in the fourth quarter of 2026 .
The Bottom Line: AI Is Driving Real Business Value
DBS has been investing in AI for more than a decade and increasingly sees the technology as a strategic differentiator rather than simply a customer service tool .
In 2025, DBS generated S$1 billion in economic value through the use of AI, a one-third increase from that generated in 2024 .
CEO Tan Su Shan told CNA that AI is beginning to contribute to fee income growth, with the technology helping relationship managers generate strategic ideas for clients and increasing transaction activity .
“Young RMs today who are AI-enabled can now go and talk to the CEO or chairman or CFO of the company and come up with great ideation on M&A or strategic ideas,” she said .
FAQ
Q: How many corporate clients now have access to DBS’s agentic AI?
A: 350,000 corporate clients in Singapore, including SMEs, can now use DBS Joy’s agentic capabilities .
Q: What can DBS Joy actually do?
A: DBS Joy can retrieve and analyse account data, answer natural-language questions about payments and transactions, and execute selected banking requests within a single conversation .
Q: Is the AI making decisions on its own?
A: No. DBS requires human oversight for all AI outputs. The relationship manager interrogates the system through a chat interface, asking it to justify its conclusions and produce supporting evidence .
Q: Will this expand to other countries?
A: Yes. The corporate agentic AI capabilities will be extended to 100,000 users in Hong Kong in September, then rolled out to other key markets .
Q: How much value has DBS generated from AI?
A: DBS generated S$1 billion in economic value through AI in 2025, a one-third increase from 2024 .
Q: How many clients does DBS Bank serve with agentic AI?
A: 350,000.
Final Thoughts
DBS’s rollout of agentic AI to 350,000 corporate clients marks a significant moment in banking. The shift from “helping customers find information” to “helping them get things done” is exactly what the next generation of AI promises.
The fact that DBS is holding back on full autonomy — demanding human oversight because governance tools lag capability — is worth paying attention to. It suggests the technology is moving faster than the safeguards, a theme likely to play out across industries.
What’s clear: AI is no longer just about answering questions. It’s about taking action. And DBS is leading the way.
DBS Bank is leading the way in agentic AI for corporate banking. DBS Bank’s agentic AI rollout is a signal that banking is entering a new era. DBS Bank’s agentic AI rollout marks a significant milestone for AI adoption in banking.
Related Posts on Pixelaizone
- HSBC Just Hired 100 AI Specialists — Here’s What They’ll Be Doing (Game-Changing Move)
- AI Ethics Explained: Why 90% of Americans Want AI Regulated in 2026 Essential Guide
- Agentic AI vs Generative AI: The Ultimate Guide to What’s the Difference and Why 2026 Matters
- “The Rise of AI Agents: 5 Powerful Tools That Actually Do Work for You”
- 11 AI Coding Tools You Should Use in 2026 (Beyond GitHub Copilot) – Game-Changing Guide
- The Best AI Image Generator in 2026: Midjourney vs Kling AI vs Adobe Firefly
- 5 Free AI Tools That Are Better Than ChatGPT for Specific Tasks
- Perplexity vs ChatGPT: Why Researchers Are Switching to Perplexity in 2026
- Claude vs ChatGPT : Which AI Tool Is Better for Long-Form Content in 2026?
- Google AI Mode vs ChatGPT: Which Search Engine Will Win in 2026? Complete Analysis
What do you think about banks using agentic AI? Drop a comment below!
