To say that William Shakespeare was a savant rather than just an extraordinary playwright wouldn’t be an overexaggeration. After all, with his words he very eloquently encapsulated some very fundamental truths about human nature, like, for example, the need for us to make a choice between, say, two options. The Macbethian dilemma of “to be or not to be” chases us always, and nowhere else is it more true than in the case of AI, where the ‘human vs. machine’ debate seems to have rather picked up some buzz in recent times. Especially if one looks at the corridors of modern banking, a question seems to have gripped almost everyone’s attention: should AI be generative or cognitive? Should it create, or should it comprehend?
To me, this isn’t a race for capability and one-upmanship, but rather a clarion call for convergence. In my view, banking doesn’t need machines that just think faster; it needs systems that understand deeper.
Generative and Cognitive AI: Two Sides of the Same Coin
Generative AI and cognitive AI are not opposing ideologies but paradigms of the same aspect in my eyes; one interprets, while the other imagines. Generative AI can compose risk models, simulate market conditions, or produce tailored credit strategies, but it’s the role of cognitive AI to lend meaning behind these outcomes, grounding creativity and process with judgement when needed.
When they function in perfect sync, banks stop merely predicting the future; they begin to own it.
The Case for Convergence
India is a land of digital access and success. It’s also the lynchpin behind how inclusive our financial landscape has truly become in recent years, with nearly 600 million consumers engaging with a wide variety of products, processes and services.
In such a scenario, the distinction between data and understanding becomes purely academic. Your LLM can generate endless insights, but it’s cognition that gives it real context, ensuring that decision-making is grounded in real empathy.
It is this interrelatedness between the cognitive and the generative that, when leveraged properly, can truly redefine key areas like risk assessment, fraud analytics, and financial inclusion.
Generative algorithms can design flexible credit lines for micro-entrepreneurs, while cognitive perspectives interpret repayment behaviour, linguistic cues, and market volatility to fine-tune risk.
The result is not automation, it’s augmentation. Banking, in this new paradigm, becomes more human because its intelligence is now more complete.
From Efficiency to Empathy
At its very core, banking has an essential human side, one that facilitates progress and empowers happiness and growth for individuals across the spectrum. Here change can’t be stymied by obtuse preferences but must be in lock and step with the evolving aspirational dynamics, and that is where the role of understanding and empathy becomes tremendously important.
I envision AI methodologies and systems that do not just manage data but perceive context, judging each case or situation by its own merits and discarding a one-size-fits-all approach.
Thus a farmer in Vidharbha might have genuinely different needs and circumstances from a senior professional in Mumbai, and AI architecture that blends generative with cognitive can truly discern that, using tools like psychometric analytics and behavioural modelling to assess risk, detect stress and evaluate with real fairness and that genuine human touch.
This transforms AI from a mere gatekeeper to an organic and proactive care mechanism, one capable of recognising patterns and emotions and aligning accordingly.
And as I say, “When technology starts listening instead of just watching, banking becomes not only smarter but also kinder and truly personal.”
Rewiring Decision-Making
For decades, banking decisions have balanced numbers against instinct. The integration of cognitive and generative intelligence makes this more explicit and integrated.
Thus, while generative models simulate macroeconomic futures and curate credit frameworks, cognitive systems bring context to it all in the form of regulatory nuance, ethical oversight, and interpretive clarity.
Together, they form what I call a “loop of reflective intelligence”, a design that learns, reasons, self-corrects, and automatically self-evolves as well.
This isn’t about automation competing with human analysis, but both working in sync, freeing up space for the human mind to decode, decipher and redefine better.
Now decision-makers and implementers get the room to exercise moral judgement, build trust, and lead with purpose.
To me, this convergence doesn’t erase the human; it just elevates it to a whole new plane. So there is no real need to embrace this with a Macbethian spirit of opposition.
Towards a Wiser Intelligence
The future of AI in banking, then, is neither purely generative nor purely cognitive; it is an integrated, interpretive, and autonomous mixture of both.
For me, the challenge before leaders is philosophical as much as technical: to ensure that intelligence remains tethered to intent.
If banking is to evolve from efficiency to empathy, from automation to awareness, then AI must become both the algorithm and the alchemist — capable of creating, understanding, and caring.