The National Payments Corporation of India (NPCI) is building a Unified Agent Protocol (UAP) to allow AI assistants to execute UPI payments. This framework aims to let verified AI agents complete transactions on behalf of users under pre-authorized limits. The move is designed to bridge the current gap where AI can plan purchases but cannot finalize the payment, potentially increasing digital transaction efficiency.
The National Payments Corporation of India (NPCI) is working on a new framework called the Unified Agent Protocol (UAP) that seeks to integrate artificial intelligence directly into the UPI payment ecosystem. At present, while AI tools can assist users in searching for products, comparing prices, or managing travel bookings, they lack the authorization to complete the actual payment. Users currently have to manually step in to finish every transaction, which breaks the seamless experience that AI promises to deliver.
Enabling AI-Driven Transactions
The UAP is designed to act as a secure verification layer that sits on top of the existing UPI infrastructure. Its primary goal is to empower registered and authorized AI agents to initiate payments after receiving specific user instructions. Instead of replacing the current system, the protocol aims to work alongside it, ensuring that transactions remain within the established security standards of UPI. For an AI agent to execute a payment, it would need to pass through this verification layer, which confirms the agent’s legitimacy and the existence of explicit user consent.
Balancing Automation and Security
One of the most critical aspects of the UAP is the implementation of spending controls. NPCI intends to structure this protocol similarly to the existing UPI AutoPay feature, where users set predefined limits for automated transactions. This means an AI agent would not have unrestricted access to a user’s bank account. Instead, the agent would only be able to initiate payments within the financial boundaries set by the user.
While this automation could streamline e-commerce and utility payments, it introduces new operational risks that the industry will need to monitor. The potential for unauthorized transaction attempts, the emergence of malicious AI agents, and the complexity of fixing liability in case of errors are factors that will likely remain under the strict oversight of the Reserve Bank of India (RBI).
The next step for this initiative will be the formalization of the framework's security architecture and the establishment of guidelines for how AI service providers can register as authorized agents. Investors and industry participants will be tracking the final rollout details, as this could influence how fintech companies and e-commerce platforms integrate AI-driven checkout experiences into their consumer apps.