SwedenTravel and aviation2020 – ongoing

Scandinavian Airlines

EuroBonus loyalty fraud detection

SAS detected 300% more small-scale loyalty fraud transactions with machine learning

WorkingCompany's own claimVendor ↗

Problem

SAS often identified loyalty fraud after scammers had traveled or stolen points. It wanted better detection without continually expanding its fraud team.

Approach

SAS deployed an Azure Machine Learning model using account, payment and travel patterns. Automated retraining kept it updated. Investigators reviewed explanations and made final determinations.

Results

300% moresmall-scale fraudulent transactions detected, per MicrosoftCompany's own claim[1]

Takeaway

SAS found more small-scale fraud while retaining human decisions on suspected cases.

Sources

Scandinavian Airlines reduces loyalty program fraud with Microsoft Azure Machine Learning
Microsoft13 May 2020Vendor publication

Source summaries

[1]Microsoft13 May 2020

Microsoft described SAS's operational fraud detection model. Its analytics and AI head reported finding 300% more small-scale fraudulent transactions.

Full details are in the original sources.

Added 4 Nov 2026 · Updated 4 Nov 2026