Payment Gateways Score Each Transaction in 100ms
Verifone's gateways now run Visa's behavioral fraud models in production, with no independent accuracy results yet published.
Updated on Nov. 4, 2026 in Artificial Intelligence
Live Poll
Do you trust AI systems to monitor your card payments for fraud?
Verifone has deployed Visa's Featurespace platform and adaptive machine learning models across its payment gateways, providing real-time assessments of each payment. The deployment was announced November 4, 2026; it is a shipped, deployed capability rather than a research-stage one.
Why it matters
Fraud detection is becoming harder as AI gives more people access to advanced attack techniques, pushing payment processors toward real-time models that combine many signals at once. Embedding these models in gateway infrastructure moves fraud scoring from a downstream review step into the transaction path itself.
The risk scoring framework evaluates more than 1,800 attributes of each transaction and keeps request processing time under 100 milliseconds, according to Verifone's specifications. The source does not provide detection accuracy, false-positive rates, or fraud-loss comparisons against prior systems, so the speed and attribute figures are the only measured specifications available.
The players
Verifone
A payments technology company that provides payment gateway and terminal infrastructure to merchants worldwide.
Visa
A global payment network whose Adaptive Behavioral Analytics and Featurespace platform now run within Verifone's gateways.
The details
Verifone's framework builds a behavioral baseline for each merchant, card, and account using Visa's Adaptive Behavioral Analytics, which records what normal activity looks like for each. When a new payment arrives, the system compares its more than 1,800 attributes against that baseline and flags patterns that appear only when several signals occur together. Because this comparison must finish within 100 milliseconds, the scoring runs as part of the payment request rather than as a later review.
Timeline
November 4, 2026: Verifone announced the Featurespace deployment across its gateways.
The Tech Race
The deployment places Visa's behavioral analytics inside a third-party gateway, extending a network-level fraud approach into merchant-facing infrastructure. Competing payment processors are likely to face the same pressure to embed real-time behavioral scoring in their own gateways.
Merchants and cardholders using Verifone-powered gateways are the parties whose transactions will pass through the new scoring, though the source gives no rollout dates by merchant or region. Availability for individual merchants, pricing, and any change to decline rates have not been announced.
The takeaway
Fraud scoring is shifting from after-the-fact review to real-time evaluation inside the payment request itself. Watch for independent fraud-loss and false-positive figures from Verifone or Visa, which would show whether the 1,800-attribute, sub-100-millisecond design delivers measurable gains.
Further reading
For broader context on how machine learning is being applied to transaction security, see the Artificial Intelligence section.
Source note: This article includes information reported by PYMNTS.
Live Poll
Do you trust AI systems to monitor your card payments for fraud?
