The sheer pace of AI developments has been breathtaking.
AI is no longer just a smart assistant. The slew of AI innovations coming on stream has been relentless. For me, this has been nothing short of amazing, having seen this firsthand in a recent trip to Silicon Valley to visit key partners and players.
OpenAI, Anthropic, Google, Microsoft, Cursor, LangChain… both big tech companies and emerging players are pushing the AI frontier at breakneck speed. Development teams are now using their platforms and bots as semi-autonomous coding agents, turning ideas into software. The concept demonstrated by OpenClaw is now mainstream as agents in Claude and Cursor can plan, code, debug and manage the full software development lifecycle. Google’s DeepMind demonstration of 3D video conferencing, songwriting and simulating the physical world with world models are fascinating.
This new reality changes what we thought would one day happen; it is happening now, unlocking a whole new world full of possibilities.
As Chief Operating Officer, my focus has been to ensure the smooth running of the bank and its transformation. Managing change and transforming with AI have been key parts of our engine and DNA for well over a decade to derive data-led insights, improve customer journeys, remove employee toil and strengthen business resiliency. Having seen what I saw, I am glad that we did not enter the AI era cold.
These AI developments have deep implications for not just DBS or banks but every industry. In 2014, our then CEO led a rallying cry for the bank to “Digitise or Die” in the face of the fintech challenge. We not only focused on data and machine learning but together forged a work culture of agility and teamwork in an approach we call “Managing through Journeys (MtJs)” that DBS is now known for. Through MtJs, we optimised how our teams collaborate at scale using a data-driven approach.
Gen AI has now given us the opportunity to redesign our operating model by leveraging human-AI collaboration. While MtJs have changed our way of working, our focus at DBS is more about Operating Model Transformation or OMTs to further unlock capacity and create greater value to fuel our growth agenda.
Our data and AI foundation developed over the past decade has allowed us to systematically re-engineer processes and workflows, upskill and reskill our people and reorganise our teams to operate alongside AI agents.
Today, over 70% of our staff utilise our in-house DBS-GPT to aid their work, with many others creating personal and team agents to leverage this capability. Vertical Gen AI use cases, including AI assistants and chatbots for customer servicing, coding agents for tech development, testing and deployment as well as the use of documentary AI, are now deeply embedded in our workflows.
It has been a great journey thus far, and having been through this, I have often been asked to share some insights with practitioners.
First, we need to get our governance right. Many users today tout the benefits agents bring to workflows. However, few can clearly explain the framework to be put in place to govern these agents across the enterprise. Imagine the deployment of multiple autonomous agents executing multi-step workflows without proper monitoring, observability and traceability. To manage this, organisations should consider putting in place an agentic control plane to ensure rigorous oversight of all agents deployed across the enterprise. The availability of tools to do this is nascent and developing, hence the need for attention.
Second, we cannot let up on our data fundamentals. In fact, the use of AI and agentic solutions requires us to have higher data quality and security. AI outcomes are only as reliable as the data consumed by AI agents. Principally, data teams will need to take centre stage to drive data quality by reducing data duplication and enforcing good metadata practices and lineage.
Third, deploying AI with a heart is a key theme that our CEO Su Shan continues to exhort – we save people, not jobs. Ultimately, AI must be a tool that works for our people. While public debate often centres on how AI changes jobs and the potential to drive higher unemployment, we need to keep our focus on the individuals behind them. Therefore, it is essential to equip our workforce with the fluency to use AI tools effectively and to reskill them for new roles sooner rather than later.
It can feel scary when the ground and everything seem to be shifting every day by AI innovations. However, AI can be a resounding win for employees, customers and the organisation – we just need to get our basics right and steer the change carefully. While there are risks in this journey, if properly managed, there is a pot of gold to be found at the end of the rainbow, for everyone.
