Pine Labs and peers test AI payment agents, but monetisation remains unproven
Pine Labs has rolled out its P3P agentic payment protocol to select merchants as Paytm, Razorpay, Cashfree and Juspay also deploy AI for payments, support and cart recovery. The near-term case is productivity and margin improvement; direct revenue models largely remain in beta amid rising infrastructure costs.
What happened
Indian fintechs including Pine Labs, Paytm, Razorpay, Cashfree and Juspay are rolling out AI agents for merchant payments, cart recovery and support. Firms
Key facts
- June: Pine Labs rolled out P3P agentic payment protocol for select merchants
- 90% of new code in at least four Pine Labs divisions is written using AI
- Q1FY26
Why this matters
Prioritise partnerships or acquisitions that add merchant-distribution, payment orchestration or proprietary transaction data, as standalone agentic-payment capabilities remain commercially unproven.
What to watch
- Pine Labs, Paytm, Razorpay, Cashfree or Juspay announcing paid agentic-payment tiers, outcome-based pricing or material pilot-to-production conversions.
- Merchant disclosures showing measurable gains in checkout conversion, payment success rates, cart recovery or support-cost reduction.
- RBI, NPCI, card-network or bank guidance on delegated payment authority, AI-agent authentication, consent and liability.
- Changes in fraud, chargeback, failed-payment or dispute rates for agent-assisted transactions.
- Evidence that inference and orchestration costs decline enough to support use at SMB transaction volumes.
- Large commerce, POS or bank partnerships that make one protocol a de facto distribution standard.
- Pilot agentic checkout and cart-recovery workflows with high-volume enterprise merchants, using human approval or policy guardrails for payment execution.
- Bundle AI support, reconciliation and conversion tools into existing merchant plans rather than launch standalone pricing immediately.
- Measure incremental authorisation rate, checkout conversion, recovery rate, support-ticket deflection, fraud losses and inference cost per transaction.
- Seek integrations with banks, card networks, UPI participants and commerce platforms to establish trusted payment-agent permissions.
- Use early deployment data to segment merchants: enterprise clients may support performance pricing, while SMBs receive AI features as retention tools.