Stripe Launches AI-Powered Fraud Detection System for Global Commerce

The payments giant introduces a revolutionary machine learning system that reduces payment fraud by 89%, processing over $1 trillion in transactions annually across 190 countries.

Fintech · North America · 2026-02-27 · 2 min read · By John Awab

Stripe has launched its next-generation AI-powered fraud detection system, Radar Ultra, which the company claims can identify and prevent fraudulent transactions with 89% greater accuracy than its previous system. The platform processes over $1 trillion in annual transaction volume across 190 countries.

The system employs a novel approach combining graph neural networks with real-time behavioral analysis, examining over 1,000 data points per transaction in under 50 milliseconds. This allows merchants to maintain seamless checkout experiences while significantly reducing chargebacks and fraudulent payments.

Patrick Collison, CEO of Stripe, described the launch as "the most significant advancement in payment security since the introduction of 3D Secure." The system is immediately available to all Stripe merchants at no additional cost, reinforcing the company's strategy of building security directly into its payment infrastructure.

Early adopters report dramatic improvements in fraud prevention. Shopify merchants using the system saw an average 73% reduction in fraudulent transactions during the beta period, while legitimate transaction approval rates increased by 12%.

The technology also includes advanced money laundering detection capabilities, helping businesses comply with increasingly stringent anti-money laundering regulations across global markets.