The company explained that this protocol tackles a significant challenge in emerging “agentic commerce” models, where AI agents can explore products, compare choices, and make purchasing decisions on behalf of users, yet face obstacles completing payments due to the manual authentication required by UPI transactions, which usually necessitates a PIN or an approval step.
Pine Labs stated that P3P enhances UPI’s current mandate-based payment framework, allowing users to pre-authorize transactions.
Once a mandate is granted, an AI agent can execute transactions within specified limits and conditions without further user involvement at the time of purchase.
This protocol integrates UPI mandate infrastructure with identity verification and authorization controls provided by Grantex, along with HTTP 402, an open web standard for machine-readable payment requests. Pine Labs noted that the system incorporates safeguards such as spending thresholds, audit trails, and users’ ability to amend or cancel permissions.
The digital gold savings platform Gullak has become the first to deploy the protocol in a live setting. Through this implementation, users can set stipulations like purchasing gold when prices drop below a certain point, allowing an AI agent to automatically carry out the transaction if the conditions are satisfied.
Meanwhile, electronics retailer Vijay Sales is testing a proof-of-concept for the technology, investigating scenarios where AI agents could autonomously acquire items when they hit a target price established by customers.
This launch aligns with the efforts of AI-driven commerce platforms to find ways to automate comprehensive purchasing experiences. While AI agents can already aid consumers in product exploration and decision-making, the execution of payments has remained a considerable obstacle due to the authentication requirements inherent in current payment systems.
Pine Labs confirmed that P3P is now operational on UPI and is collaborating with card networks to extend the protocol to card-based transactions in the future.
First Published: Jun 12, 2026 7:49 AM IST