Tuesday, October 27, 2026 | 12:00 PM ISTहिंदी में पढें
Business Standard
Notification Icon
userprofile IconSearch

Agentic payments need clear consent and security rules: Industry leaders

At the BFSI Summit 2026, industry experts said AI agents could simplify everyday payments, but adoption hinges on verifying their identities and permissions

BFSI summit, Nirav Kamdar, CTO, Amazon Pay, Pramod Mulani, VP Global Strategy, Juspay and Rishi Chhabra, Country Manager, India, VISA

AI-driven payments are reshaping digital commerce, raising questions about security, consumer consent, and accountability.

Sarjna Rai New Delhi

Listen to This Article

 As artificial intelligence (AI) agents begin to search, shop, and make payments on behalf of consumers, clear rules on authorisation, authentication, and liability will be essential to building trust in agentic payments, industry leaders said on Day 2 of the Business Standard BFSI Insight Summit 2026 on Tuesday.
 
During a panel discussion titled 'Machines with Mandates: Building India’s Trust Layer for Agentic Payments' with Business Standard's Ajinkya Kawale, Visa India Country Manager Rishi Chhabra, Juspay Vice-President of Global Strategy Pramod Mulani, and Amazon Pay Chief Technology Officer Nirav Kamdar discussed the challenges of enabling AI agents to transact independently while keeping consumers and merchants protected.
 
 

Agentic payments: Early use cases

 
According to Mulani, travel bookings could be among the earliest use cases for agentic commerce because they involve multiple decisions, including comparing fares, selecting dates, and choosing airlines.
 
“Wherever there is a lot of variability, I think agents will add a lot of value,” Mulani said, adding that AI agents could help consumers plan itineraries, find suitable flights, and secure better deals.
 
Kamdar, however, expects routine transactions to gain traction early. “Those are the use cases where agentic will pick up. The risk is low, friction is high, so it’s a good combination,” he said, referring to utility bill payments and subscription renewals. 
 
AI agents could also monitor product prices and complete purchases when prices fall below a specified threshold. However, widespread adoption will depend on consumers becoming comfortable with delegating decisions and transactions to machines.
 

Building trust in AI agents

 
Chhabra said trust must remain central to the process, ensuring that agents act within the limits set by consumers and that transactions can be traced from beginning to end.
 
Building trust will require several safeguards, including agent authentication, identity verification, fraud detection, explicit consent, and auditability. Consumers must trust agents to act on their behalf, while merchants and platforms must be able to verify that the agents are authorised to do so.
 
Chhabra said Visa has been developing its Intelligent Commerce framework to support AI-driven shopping and payments. He shared the company had also introduced a Trusted Agent Protocol to help merchants verify an agent’s identity and determine what actions it is permitted to perform.
 

Need for common framework

 
Kamdar said the industry should not assume that a single protocol would dominate. Instead, financial institutions and fintech companies could build common capabilities for discovering, identifying, and authenticating agents, alongside governance and auditing safeguards, while adapting to different protocols.
 
He said separating the intelligence layer from the protocol layer could help the industry develop systems more quickly without requiring every platform to adopt the same technology.
 
Mulani said common frameworks would likely emerge, much as they have for conventional card payments and online transactions. Individual payment networks could then adapt these frameworks to their own systems, while country-specific regulations would continue to shape implementation.
 

The liability challenge

 
Mulani said the industry was still establishing what information must travel across the payment ecosystem to verify consumer intent and determine whether an agent acted within its authorisation. Clearer liability rules would follow once these foundations were established.
 
Chhabra said the industry would need to ensure transactions were traceable, from the consumer’s original instructions to the agent’s actions, before responsibility could be assigned when something went wrong.
 
Kamdar stressed that a robust liability framework would be critical to consumer confidence. Drawing a parallel with road safety, he said, “It’s not the accelerator; it’s the fact that there is a brake. That’s what gives you the confidence that you can move fast.”  

Don't miss the most important news and views of the day. Get them on our Telegram channel

First Published: Oct 27 2026 | 11:46 AM IST