Analysis · Financial Services

Kiwi cofounder says AI assists humans in Indian banking

By Startup Enthusiast · 25 February 2026

What they do

Siddharth Mehta, cofounder of Kiwi, says AI acts as an enabler that uses layered checks and escalation steps, not full automation

AI in India's BFSI sector currently helps with fraud prevention, risk scoring, document checks, customer service, and internal tasks

Core money moves and credit choices stay under human watch because the industry needs explainable and auditable AI systems

Siddharth Mehta, cofounder of Kiwi, states that AI acts as an enabler rather than an autopilot in the Indian banking sector.
AI applications in India's BFSI ecosystem include fraud prevention, risk scoring, document verification, customer servicing, and internal productivity.
Core money movement and credit decisions remain under strict human oversight to protect capital and ensure accountability.
The industry expects explainable, auditable, and accountable AI stacks before deeper automated decision roles are adopted.
Data foundations, governance frameworks, and oversight mechanisms must mature before AI can take on more significant decision-making responsibilities.
Payments providers are reluctant to give AI full control due to asymmetric risks involving trust damage and regulatory implications.
Axis Bank has restructured its legacy data organization to build curated data layers and semantic frameworks for AI-readable structured banking data.
Axis Bank achieved a traditional multi-year transformation in 12-15 months through disciplined AI implementation and human-in-the-loop oversight.

Ecosystem maturity, not algorithmic capability, is identified as the primary bottleneck for enterprise-wide AI deployment in Indian finance.
Separate data systems and weak links between institutions make it hard and expensive to roll out AI across the whole industry.
Banks are holding back from putting AI in charge of big financial choices until common rules for fairness and model checks are in place.

Kiwi monitors model accuracy, false positives, and behavioral drift continuously to ensure reliable AI performance.
Plutos ONE transitioned to an AI-first organization while anchoring innovation in secure, auditable, and regulator-ready systems.
Cashfree Payments focuses on responsible AI deployment to strengthen product value and sustain customer trust.
PayU offers an AI payments solution for hosted checkouts that keeps the user involved and meets compliance rules.
CoverSure treats governance as a design principle rather than a final checkpoint to ensure compliance cannot be retrofitted later.
Oriserve identifies infrastructure and data silos as primary barriers preventing AI from moving into decision-grade roles in core banking.
Most banks operate decades-old core systems that were not designed for seamless AI integration or standardization.
Pilot projects succeed in areas like collections and lead qualification but face constraints in enterprise scaling due to integration issues.

Product: An app that issues RuPay credit cards and offers credit lines linked to the UPI network.
Customers: Individual consumers seeking the convenience of UPI payments combined with credit card limits.
How it makes money: Partners with banks to issue co-branded credit cards and credit lines, earning transaction-based fees.
What sets it apart: Pioneered Credit-on-UPI, allowing users to make credit-backed payments at any UPI-accepting merchant.