Visa outlines AI, agentic commerce and cyber resilience roadmap for payments
Visa India & South Asia’s Suresh Sethi highlighted agent-ready merchant websites, secure AI-led transactions and stronger cybersecurity as priorities for India’s expanding digital-payments ecosystem.
What happened
Visa India & South Asia’s Suresh Sethi outlined how AI, agentic commerce and cybersecurity will shape digital payments, highlighting trusted infrastructure,
Why this matters
Payments, cybersecurity and commerce-platform targets with agent-ready capabilities could become strategically valuable as retailers and networks build secure AI-led transaction ecosystems.
What to watch
- Visa launches India-specific agentic-commerce APIs, token standards, delegated-authorization products or issuer/acquirer pilot programs.
- Major Indian marketplaces, quick-commerce firms, travel platforms or omnichannel retailers announce AI agents capable of completing payment rather than only recommending products.
- RBI, NPCI or card-network guidance on AI-agent authorization, consent, data sharing, chargeback liability and transaction limits.
- Measured changes in card tokenization, credential-on-file approval rates, fraud losses and friendly-fraud disputes at major merchants.
- Cyber incidents involving AI shopping assistants, merchant API manipulation, synthetic identities or unauthorized agent-led purchases.
- UPI adds comparable agent-payment permissions or interoperable recurring/delegated-payment features, increasing competitive pressure on card networks.
- Prioritize tokenized credential-on-file, passkey, device-binding and delegated-authentication capabilities before enabling agent-initiated checkout.
- Build structured, agent-readable merchant catalogs covering price, availability, delivery promises, returns and loyalty eligibility; inaccurate feeds will become a conversion and dispute risk.
- Create distinct agent-payment controls: customer-set spending limits, merchant allowlists, category restrictions, step-up verification and real-time transaction alerts.
- Treat fraud, chargebacks and consumer-redress workflows as product requirements; assign liability rules for agent errors before scaling pilots.
- Partner with issuers, acquirers and payment networks on transaction-risk scoring that distinguishes human, merchant-agent and consumer-agent activity.
- Use cross-border, travel, premium retail and subscription use cases as early agentic-payment pilots where card credentials and fraud tools have clearer advantages over domestic instant-payment rails.