A new-generation intelligence framework for banking
KYC to UYC, AI to MI to HI to Total Intelligence
Can All These Bring Ethics and Accountability to Digital Payments?
Digital payments, particularly through UPI, have grown by leaps and bounds, reaching levels that were almost unimaginable a few years ago. In large parts of the country, especially in the retail and mass-payment segments, the use of currency notes has declined substantially. This is a remarkable reflection of India's technological progress and, more importantly, the widespread acceptance of that progress by its people.
The growth of digital payments has enormous potential for the economy. It can improve transparency, reduce transaction costs, widen financial inclusion and gradually overcome several weaknesses associated with cash-based transactions and traditional administrative systems.
But technology, however advanced, is only a means. The more important question is whether technology can also help strengthen ethics, values, honesty, integrity and accountability in the financial system.provided technology is combined with human judgement and a deeper understanding of the customer. From “Know Your Customer” to “Understand Your Customer”
The banking system has long operated on the principle of Know Your Customer (KYC). KYC is indispensable for establishing the identity and address of a customer and preventing the misuse of banking channels. But in the digital era, knowing who the customer is may no longer be enough. What may increasingly be required is “Understand Your Customer” (UYC).
Within the limits of privacy and legitimate banking practice, UYC could help banks understand a customer's normal transaction behaviour, financial patterns, likely vulnerabilities and significant deviations from established patterns. Artificial Intelligence (AI) can make such analysis possible on a scale that human beings alone cannot manage. This could help identify suspicious transactions, unusual transfers, possible account misuse and even genuine mistakes before they become difficult to rectify.
A Simple Example
Consider a customer of Bank A making a UPI payment through a third-party interface such as PhonePe and, through an inadvertent mistake, transferring money to an unintended beneficiary whose account is with Bank B. The customer realises the mistake and immediately approaches Bank A. Technically, the payment was authorised by the customer and the money has already moved through the payment system. The payer's bank may therefore have limited ability to reverse the transaction unilaterally. But should the matter necessarily end there?
Once the error is reported, the banking and payment ecosystem should, wherever legally permissible, have mechanisms to identify the transaction, alert the beneficiary bank, place an appropriate hold or caution and facilitate recovery when the circumstances establish that the payment was genuinely erroneous.
Technical authorisation should not, by itself, settle the larger question of ethical ownership and institutional responsibility. The beneficiary has received money which, in such circumstances, was never intended for him or her. The banking system should therefore endeavour to facilitate correction rather than simply treating the transaction as closed because authentication was successful.
The Reserve Bank of India, perhaps through appropriate mechanisms under its Banking Ombudsman framework, together with banks' own persuasion and moral-suasion approaches within permissible parameters, could consider workable arrangements for dealing with such cases. The objective should be simple: to help ensure that money ultimately reaches its rightful destination based on entitlement and legitimate obligation. The person not entitled or intended by the sender cannot be allowed to enjoy such wrong credits by any means. Human or technical error should not be allowed to override ethics and morality.
AI as a Second Line of Defence .
This is where AI can become a powerful second line of defence. AI can potentially detect unusual behaviour at the point of transaction—a sudden departure from established payment patterns, unusually large or repeated transfers, unusual beneficiary relationships or other warning signals. Depending on the risk involved and the regulatory framework, such signals could trigger an alert, confirmation or additional verification. More importantly, AI need not stop working once the transaction is completed.
In the post-transaction stage, technology could help trace the movement of funds, identify the beneficiary institution, establish the relevant circumstances and facilitate a structured resolution when a genuine mistake has occurred. Technology should therefore not merely make money move faster. It should also make mistakes easier to detect and rectify.
Fraud Prevention Requires More Than KYC . Frauds involving bank accounts, identity misuse, social engineering and exploitation of digital platforms are constantly evolving. KYC establishes who the customer is, but it does not necessarily establish whether a particular transaction is consistent with that customer's normal behaviour. That is the gap UYC could help address.At the same time, such an approach must never become an excuse for indiscriminate surveillance or intrusion into legitimate customer privacy. There must be clear safeguards, proportionality, transparency and accountability in the collection, analysis and use of customer information. The objective should be better understanding, not unnecessary intrusion.
AI Cannot Replace Human Intelligence .There is also a danger in believing that AI can solve everything. It cannot. AI can identify patterns, correlations and anomalies at extraordinary speed. But Human Intelligence (HI) provides judgement, context, experience, empathy and accountability.The objective should therefore not be AI replacing human beings, but AI strengthening human intelligence and institutional responsibility. A genuinely customer-oriented bank should be able to ask not merely:
“Was the transaction technically authorised?”but also “Does the transaction make sense in the context of the customer, and if something has clearly gone wrong, what responsibility can the banking system reasonably assume in helping to correct it?” That represents an important change in institutional thinking—from merely processing transactions to understanding their circumstances and consequences.
Market Intelligence Adds Another Dimension
There is yet another dimension that can make this framework much more comprehensive: Market Intelligence (MI). KYC tells the bank who the customer is. UYC helps it understand how the customer normally behaves. AI analyses patterns and anomalies, while Human Intelligence interprets what those signals actually mean. But a customer does not operate in isolation.
Market Intelligence can provide information about the environment in which the customer operates—industry conditions, markets, competitors, suppliers, customers, prices, demand, regulatory developments and emerging risks.
A Powerful Application: Credit and NPA Prevention
This becomes particularly valuable in credit appraisal and NPA prevention. A borrower may have an excellent KYC profile and an apparently satisfactory account. Yet Market Intelligence may reveal that the borrower's industry is deteriorating, major customers are under stress, commodity prices have changed sharply, or the market itself is becoming increasingly unviable.
AI may simultaneously detect corresponding changes in the borrower's account operations. Human Intelligence can investigate and interpret the signals. The bank can then take appropriate corrective action before the problem develops into an NPA. The same approach could strengthen monitoring of the end-use of funds, early-warning systems and responsible lending. From Identification to Responsible Decision-Making.Thus, a much broader intelligence architecture begins to emerge:
KYC + UYC + MI + AI + HI = Total Intelligence
Each has a distinct role:
KYC identifies.
UYC understands.
MI contextualises.
AI analyses.
HI interprets.
Total Intelligence decides.
The concept is not to create a system that knows everything about everyone. Nor is it intended to create excessive surveillance. Its purpose should be to enable banks to make better, more responsible and timely decisions while respecting privacy, legitimate freedom and the rights of customers.
From Transaction Processing to Responsible Banking
No financial system can realistically be made completely foolproof. Human beings will always find new ways of circumventing systems. But an intelligent combination of KYC, UYC, Market Intelligence, AI and Human Intelligence could make banking considerably more preventive, transparent, resilient and accountable. India has already demonstrated that technology can transform the way millions of people transact. UPI is a powerful example of how rapidly people can adopt technology when it is convenient, inexpensive and accessible.
The next challenge is to ensure that trust keeps pace with technologyDigital payments should not merely be fast; they should be safe They should not merely be convenient; they should be accountable.They should not merely prevent fraud; they should also provide meaningful mechanisms for correcting genuine mistakes. And banks should not merely “know” their customers; they should increasingly understand them, while respecting privacy and individual freedom. Ultimately, technology should not be allowed to become an excuse for avoiding responsibility. The more powerful our technological systems become, the greater should be the emphasis on human values, ethical judgement and institutional accountability.
The real opportunity before the banking system is therefore not simply to build a more sophisticated digital payment system, but to build a more intelligent and responsible financial system.Technology has already changed the way India moves money. The next step is to ensure that it also strengthens the values with which money is handled.That may be the true meaning of moving from KYC to UYC, from AI and MI to HI, and ultimately towards Total Intelligence.
Samastha Loka Sukhino Bhavanthu.
T V G Krishnan
(Personal Views)