Feedzai launches IQ Score fraud-risk API for banks
Feedzai launched IQ Score, an AI-native API that gives banks real-time anonymized signals from its $9 trillion transaction network without sharing customer-level data.
Feedzai announced Feedzai IQ Score, an AI-native, network-derived fraud risk scoring API that provides banks with real-time, anonymized signals drawn from the company’s $9 trillion global transaction network without exposing customer-level data.
The API delivers pre-trained AI models via a lightweight integration, removing the need for banks to supply historical data or run lengthy model training cycles. Feedzai stated institutions can move from integration to value in days and that the system requires minimal operational effort to deploy. For faster-payment use cases, as few as 15 data fields are sufficient for the solution to start producing risk signals.
IQ Score is built on a federated intelligence architecture that shares aggregated, anonymized patterns of fraud across the network while keeping raw customer records local. The company contrasted that design with consortium-style data exchanges and added it is intended to reduce the transfer of sensitive data and attendant compliance questions.
Feedzai reported detection improvements in its tests, citing four times more fraud identified and 50% fewer alerts compared with traditional rules-based systems. The product is offered as an API that can run alongside existing risk platforms.
The offering is targeted at banks of all sizes but is presented as particularly relevant to regional and mid-sized institutions that may not have the transaction volume or internal AI teams needed to build advanced fraud models. By using network-level signals, those institutions can apply aggregated intelligence to detect multi-channel and sophisticated fraud attempts without rebuilding their full risk infrastructure.
According to Pedro Barata, Feedzai’s chief product officer, “Fraud has outpaced what any single institution can stop alone.” He added the company expects IQ Score to give banks shared, network-level signals that can be used with their current systems.
Philip Mackenzie, senior research principal at Chartis, described network fraud intelligence sharing as increasingly important for tracking fragmented fraud signals across the financial ecosystem and said real-time cross-institutional insights are a key differentiator for Feedzai’s solution.
Feedzai stated IQ Score is available to institutions through an API and can be integrated alongside existing fraud detection and risk-scoring tools.







