As digital finance becomes indispensable, algorithmic convenience must not come at the cost of institutional accountability. AI generated image Last week, the National Payments Corporation of India announced a significant change to India’s Unified Payments Interface. From October 15, eligible merchant payments exceeding Rs.2,000 will attract a Merchant Discount Rate of 0.4 percent, capped at Rs.300 per transaction. The charge will be payable by merchants, not customers, while small merchants receiving up to Rs.1 lakh a month through QR-based UPI payments will be exempt. Certain transactions involving fuel, railways, telecommunications, insurance, utilities, agricultural inputs and government payments will attract a flat charge of Rs.5. The government has directed banks and payment providers not to pass the charge on to consumers. Nevertheless, the announcement has caused unease among merchants and users. Retailers fear that even a charge formally imposed on businesses could eventually be absorbed through higher prices or discourage some establishments from accepting larger UPI payments. The government and payments industry have a legitimate counterargument: a digital network of this scale cannot remain secure, reliable and innovative without a sustainable source of revenue. UPI reportedly processed around 24 billion transactions worth approximately $311 billion in August 2026 alone. Maintaining that infrastructure, strengthening cybersecurity, resolving disputes and preventing fraud involve substantial expenditure. The controversy raises a question that extends beyond the price of a transaction: as digital finance becomes an essential public utility, how should its costs, risks and responsibilities be distributed without undermining public trust? The Automation Paradox India’s digital-payments revolution is among the country’s most remarkable technological achievements. Yet, as transactions surge and the Reserve Bank of India tightens regulatory safeguards—from digital-lending rules to fraud prevention and customer protection—an operational paradox has emerged. The financial system is becoming faster and more automated, but customers who encounter errors can find it increasingly difficult to reach a person capable of correcting them. The RBI introduced the Ombudsman Scheme for Digital Transactions in 2019. In 2021, it launched the Reserve Bank–Integrated Ombudsman Scheme, combining the earlier schemes covering banks, non-banking financial companies and digital transactions. Despite this framework, complaints received through the integrated grievance-redressal system rose from 11,75,075 in 2023–24 to 13,34,244 in 2024–25, an increase of 13.55 percent. Loans and credit cards together reportedly accounted for nearly 46 percent of complaints, while digital banking, deposits and ATM-related grievances also constituted significant categories. These numbers do not mean that every complaint was maintainable or that misconduct was established in every case. Greater financial participation, rising transaction volumes and easier online filing can themselves generate more complaints. Indeed, more than 91 percent of complaints were reportedly submitted through digital channels. Rural and less digitally literate customers may also remain under-represented in formal grievance statistics. Nevertheless, more than 1.33 million complaints represent an important institutional warning. Rapid digitisation must be accompanied by equally accessible, responsive and accountable grievance resolution. In the early years of financial technology, the promise was straightforward and revolutionary: reduce the bureaucracy of traditional banking, eliminate physical queues and democratise access to financial services through code. Transactions that once required paperwork and repeated visits to a bank can now be completed within seconds. But as fintech has evolved from a disruptive innovation into the nervous system of everyday commerce, human discretion has sometimes been pushed to the margins. Frontline customer-service representatives often working through outsourced centres, automated interfaces or tightly controlled scripts may possess neither sufficient information nor the authority to examine exceptional cases. When a customer encounters an unexplained account restriction, an unsuccessful merchant settlement, an erroneous fraud alert or a missing interbank transfer, the initial response is generally automated. The customer may be told that “the system does not permit an override”, that the transaction was “processed automatically” or that the complaint has been “escalated.” Such responses may accurately reflect the limitations placed on the employee, but they do not resolve the customer’s problem. A decision affecting access to money has been made, yet no identifiable person appears authorised to reconsider it. Technology, which should assist institutional judgment, effectively becomes the final decision-maker. Right Versus Right Automated financial controls are not inherently unfair. Banks and payment platforms must process millions of transactions while detecting fraud, identity theft, money laundering and cyberattacks. A system that temporarily blocks an unusual transaction may prevent a customer from suffering a serious loss. The problem arises when a provisional automated decision becomes practically irreversible because the institution lacks a timely and meaningful system of human review. Ethicist and author Rushworth Kidder described such situations as “right versus right” dilemmas. One duty is to enforce regulatory and security controls; the other is to treat customers fairly, examine individual circumstances and restore access when a restriction proves mistaken. Both duties are valid. A responsible institution balances them through trained judgment, documented escalation and proportionate safeguards. Organisational incentives, however, may not always support that balance. If customer-service performance is measured primarily by call duration, ticket volume or speed of closure, employees may be encouraged to close complaints rather than solve them. A technically closed ticket can still represent an unresolved financial injury. This is where algorithmic convenience becomes institutional risk. Fintech platforms are not merely software businesses. They are custodians of money, information and confidence. Their relationship with customers rests not only on contractual conditions but also on an expectation that errors will be explained and corrected fairly. Traditional banking should not be romanticised. It, too, suffered from delays, bureaucracy and arbitrary conduct. But customers could ordinarily identify a branch, manager or departmental authority responsible for reviewing a disputed decision. Digital finance has removed much of the physical inconvenience; in some cases, it has also obscured the path of accountability. The UPI Test The UPI charge controversy makes this question particularly important. The proposed MDR may help finance cybersecurity, infrastructure, innovation and customer support. At 0.4 percent, it is lower than charges generally associated with debit and credit cards. The exemption for small merchants and the Rs.300 ceiling also attempts to contain its impact. But a sustainable payment system requires more than collecting charges. Institutions must explain who pays, how the revenue is shared and how it will improve reliability, fraud protection and grievance resolution. Authorities must also monitor whether merchants indirectly recover the charge from customers or return to cash for higher-value transactions. Transparency will determine whether the measure is seen as reasonable infrastructure financing or as a breach of the expectation that UPI payments would remain free. The solution is not to place unrestricted override powers in the hands of every customer-service employee. That could weaken legitimate controls and create opportunities for abuse. Instead, institutions need trained escalation teams with sufficient authority to review high-impact decisions, especially account restrictions, blocked funds and rejected legitimate transactions. Such decisions should be explainable, reviewable and time-bound. Institutions should maintain records showing why a restriction was imposed, who reviewed it and when the customer received a reasoned response. Where an institutional or technological error has occurred, the primary responsibility for investigation and correction should rest with the institution, rather than forcing the customer through an endless sequence of automated replies. Customer support must not be treated merely as a cost centre to be reduced through chatbots. Complaints can reveal weaknesses in fraud-detection models, payment architecture, software design, vendor management and internal controls. Properly analysed, they form an early-warning system for operational and reputational risk. The RBI’s Integrated Ombudsman Scheme provides an external remedy when an eligible complaint against a regulated entity is not resolved satisfactorily through the prescribed internal process. But the Ombudsman should be the final avenue and not the customer’s first realistic opportunity to encounter independent human judgment. (The writer is an advocate and a public policy analyst. Views personal.)
Comments