Identity Anomaly Detection

Scores login activity and security signals to spot suspicious identity behavior and trigger extra authentication or investigation.

Business impact

• Account login fraud — American Express says the model helps reduce account login fraud.
• Intrusion detection — Hugging Face’s anomaly detection flagged the intrusion.

Data requirements

• Login behavior (tabular) — Used to score account logins and detect unusual patterns.
• Feedback (tabular) — Used to calibrate risk in the login model.
• Security signals (tabular) — Used to separate meaningful events from background noise.

AI methods and techniques

• Machine learning risk scoring — Scores account logins in real time based on login behavior and feedback.
• Anomaly detection — Flags unusual signals and helps distinguish meaningful activity from background noise.

Industries

Real-world evidence

2 documented case studies on record.

Companies using this: American Express, Hugging Face.