By Guest contributor |
The financial services sector is often characterised by its rapid pace, wealth of data and demand for high-precision decision-making. The industry has long leveraged technological advancements to enhance efficiency, from early computational models to sophisticated algorithms.
Now, agentic AI is emerging as the latest disruptor for business operations, and it promises a future where intelligent systems operate in partnership with organisations to provide unprecedented autonomy and adaptability.
This technology goes beyond passive data retrieval and traditional automation, and moves towards autonomous, goal-orientated intelligence. Agentic AI systems are defined by their ability to perceive environments, reason through complex situations, create multistep plans and adapt independently to achieve objectives with minimal human supervision.
A vast number of potential use cases for agentic AI have already been identified across the financial services landscape, with key areas including efficiency, productivity and customer experience.
For instance, an increasing number of customers now expect personalised interactions with the services they use. Agentic AI can proactively analyse vast amounts of customer data – including spending patterns, life events and financial goals – to proactively recommend products and services when they are most relevant for the customer. This will not only prompt an increase in conversion rates, but also a reduction in customer support wait times.
The financial sector is inherently exposed to a range of risks such as fraud, which is where agentic AI can also help. Agents can be deployed to continuously monitor transaction patterns in real time, perform continuous credit risk assessments, autonomously monitor financial markets and carry out other tasks.
To successfully transition from the theoretical potential of agentic AI to practical implementation, organisations need to develop a structured, strategic approach. Ideally, the deployment will be carried out in phases to minimise potential risks and observe the return on investment in different areas.
The first three to six months should be the foundation stage where the basic infrastructure is established. Organisations need to align AI initiatives with their overall business strategy, which will help them to maintain a clear vision of their digital transformation rather than deploy technology for the sake of it. During this stage, they will also need to ensure that their data infrastructure is unified, up to date and accessible so the AI model can use it effectively.
In months six to 12, organisations can launch pilot tests to demonstrate the key areas of deployment. Once organisations are in the second year of implementation, they can shift their focus towards scaling these solutions across the organisation, then concentrate on continuous improvement and innovation for the future.
It is important to note that organisations do not have to take an ‘all or nothing’ approach; a hybrid approach that blends traditional automation technology and agentic AI is just as effective. In this scenario, the traditional automation can handle predictable, repeatable tasks while agentic AI tackles the more dynamic and complex challenges. This allows for a less disruptive transition.
Regardless of which method an organisation chooses, it is important for them to remember that human oversight remains indispensable throughout the process. AI agents will not take over human activities, but will complement them by taking over some of the repetitive tasks that can be automated through AI-powered decision making. Therefore, employees should maintain oversight of critical AI decisions, especially in a high-risk sector such as financial services. Furthermore, organisations should invest in training and upskilling employees to ensure a seamless collaboration between humans and machines and prevent employees from feeling displaced.
Organisations should also consider implementing an AI centre of excellence, which will act as a centralised knowledge group that guides and oversees the development of organisation-wide AI projects.
Gartner’s October 2024 AI intelligence report states that by 2028, 33 per cent of enterprise software applications will include agentic AI, and 15 per cent of day-to-day work decisions will be made using agentic tools. This implies that financial institutions that do not embrace agentic AI risk falling behind their competitors. Consequently, the integration of agentic AI is not merely an option for organisations in the financial sector, but a strategic imperative for institutions aiming to thrive in a competitive landscape. It’s no longer a distant future but a near-term reality that demands proactive engagement.
Discover more insights like this in the Summer 2025 issue of Technology Record. Don’t miss out – subscribe for free today and get future issues delivered straight to your inbox.
Niloy Sengupta is vice president and chief technology officer of US financial services
at Kyndryl