India's NPCI is examining protocols for AI agent identification within the UPI ecosystem. This move aims to enable autonomous payments on the world's largest retail fast-payment network. The framework will allow AI agents to make small digital payments without user approval. Existing UPI capabilities will be built upon to support these agentic payments.
The National Payments Corporation of India (NPCI) is examining protocols that could allow digital AI agents to be identified and authorised within the UPI ecosystem, as India looks to bring autonomous payments to the world’s largest retail fast-payment network by transaction volume.
Ajay Kumar Choudhary, non-executive chairman and independent director of NPCI, said the payments infrastructure will need to evolve as AI agents move from assisting users to acting on their behalf.
“NPCI is also examining the protocols that may be required to identify and authorise digital agents within the UPI ecosystem while preserving interoperability, auditability and settlement finality,” Choudhary said at the Global Fintech Fest 2026.
His comments come days after Reuters reported that India was preparing a framework that would allow AI agents to make small digital payments without requiring user approval for every transaction. The report, citing three people familiar with the matter, said such a framework could make UPI one of the world’s largest networks for agentic payments.
The idea is to move payments from a traditional model of users explicitly instructing a payment to one where an AI agent understands an objective, evaluates options and executes transactions within limits authorised by the user.
For instance, an agent could be asked to buy groceries within a fixed budget, identify the best available discount and complete the transaction without seeking approval each time. More advanced applications could include agents making purchases based on sale prices or even executing investments when specified price thresholds are met, Reuters reported.
The framework is expected to build on existing UPI capabilities such as UPI Circle, which allows a primary account holder to delegate payment authority, and Reserve Pay, which enables funds to be blocked for multiple debits.
NPCI is also expected to provide infrastructure for merchants to integrate with such systems, with spending limits, identity checks and audit trails forming part of the framework. A liability mechanism would also be needed as responsibility shifts between users, agents, banks and merchants.
For NPCI, however, the bigger challenge is ensuring that the flexibility of AI does not undermine the certainty required in payments.
“An agent may interpret intent, but authoriser must verify identity, must verify mandate, must verify limits, and consent,” Choudhary said.
He said the architecture should therefore separate intent, authorisation and settlement. While an AI system could interpret what a user wants and determine how to execute it, the final settlement layer would remain rule-based, secure and legally final.
This becomes particularly important as AI agents, tokenisation and payment systems increasingly converge. AI could provide the intelligence and coordination, tokenisation could make transactions programmable, while payment rails would provide the trusted mechanism for final settlement.
The shift also introduces new risks. Choudhary warned that model risk could become payment or conduct risk if AI systems are allowed to influence financial transactions. Risk management, he said, would therefore have to follow the activity across the entire chain rather than treating each technology in isolation.
The scale of UPI makes the development particularly significant. UPI processed 24.51 billion transactions worth Rs 29.82 lakh crore in August 2026 with Google Pay and PhonePe accounting for roughly three-fourths of the monthly volume.
Payment companies globally are already working on agentic commerce, while Mastercard and Visa are developing agentic payment capabilities in India. Pine Labs has also launched an agentic protocol that allows AI agents to complete UPI payments after a single upfront authorisation.
For India, the challenge now is not simply allowing AI agents to make payments, but ensuring that autonomy remains bounded, accountable and auditable as UPI enters its second decade.
