Introducing Fraud Score: A specialized model for every type of fraud, trained on your own data

Fingerprint has launched Fraud Score, an Enterprise-plan feature for Smart Signals customers that creates separate machine-learning models for each fraud category a business tracks, such as account takeover, payment fraud, fake sign-ups, and promotion abuse. Customers provide labeled examples of confirmed fraud and legitimate activity, which Fingerprint combines with device intelligence signals to train, review, and deploy customizable models after customer approval. Each visitor receives a 0-to-1 risk score for every selected fraud type through the existing API, allowing businesses to set their own response thresholds or send the scores to fraud orchestration systems. The models retrain as new outcomes are shared, aiming to adapt to changing fraud tactics while reducing the need for internal model maintenance. Fingerprint also offers proof-of-value testing using customers’ historical fraud outcomes before commitment.

No tracked trend matches for this post yet.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.