AI in Payments
Agentic checkout, fraud models, and how AI is rewriting the transaction.
JPMorgan Deploys LLM-Based Anomaly Detection Across $10T in Daily Payment Flows
JPMorgan applies large language models to payment message analysis, catching compliance and operational anomalies across its $10 trillion daily transaction volume.
FIS Worldpay Deploys AI Authorization Optimizer That Routes Transactions to Maximize Approval Rates
Worldpay launches an AI system that dynamically routes transactions across acquiring banks and networks to maximize authorization rates, recovering $1.4B in previously declined volume.
PayPal AI Shopping Assistant Lifts Merchant Conversion 19% in Controlled A/B Tests
PayPal's AI-powered checkout assistant, which recommends payment methods and offers based on user context, improves conversion rates by 19% versus standard checkout flows.
Stripe Radar Adds Generative AI Rules Engine: Merchants Describe Fraud Patterns in Plain English
Stripe launches an AI layer on top of Radar that lets merchants create custom fraud rules by describing suspicious patterns in natural language rather than writing code.
Mastercard AI Fraud Model Blocks $2B in Fraudulent Transactions in Q1, a 300% YoY Improvement
Mastercard's next-generation AI fraud detection system blocked $2 billion in fraudulent transactions in Q1 2026, tripling its 2025 performance while cutting false declines 28%.
Anthropic and Visa Pilot Agentic Checkout: AI Agents Complete Purchases Without Human Intervention
Visa and Anthropic demonstrate AI agents that autonomously browse, compare, and purchase products using tokenized payment credentials — no human in the loop.