Page 313 - Cyber Defense eMagazine September 2025
P. 313
2. Synthetic Identities: A Quiet but Devastating Attack Vector
In the arms race of digital fraud, synthetic identity creation has become one of the most challenging
threats to detect. Fraudsters build completely fake identities using real data fragments, such as legitimate
Social Security numbers, and couple that information with forged addresses to replicate a new identity
across multiple accounts.v
As document fraud becomes increasingly sophisticated, relying on static template matching is no longer
viable. Security teams must adopt document liveness models that can detect PII data manipulation. This
includes the absence of security features like holograms, photocopies, digital copies, and superimposition
of the photo area to establish document validity.
3. Uncovering organised fraud rings
A pattern we observe typically with fraud rings is that ID & Selfie images in most cases have similar
backgrounds. This visual uniformity is often overlooked by traditional ID verification systems, allowing
multiple fraudulent identities to slip through as seemingly legitimate.
Networked AI models, which evaluate the broader behavior of user identities across identity transactions,
devices, and IPs , can help detect patterns indicative of synthetic identity networks. By combining cross
transactional risk signals, identity intelligence and document forensics, organizations can uncover
clusters of fraud that may otherwise go unnoticed. This includes detecting both organized fraud rings
operating across multiple digital platforms and individual users attempting scams on multiple fronts.
Fighting Fraud with AI
Traditional verification methods, even when automated, can’t stand alone against today’s AI-enhanced
threats. Security teams must adopt a layered approach powered by advanced identity intelligence.
This means embracing biometrics-based security tactics like liveness detection to combat camera
injection and deepfakes, and incorporating adaptive, cross-transactional risk signals to expose synthetic
identity networks.
With AI working on our side, not just against us, we can build adaptive systems that meet fraudsters at
their level and stay one step ahead.
Cyber Defense eMagazine – September 2025 Edition 313
Copyright © 2025, Cyber Defense Magazine. All rights reserved worldwide.