Skip to content

Synthetic Identity Fraud

Definition

Synthetic identity fraud creates a new, fabricated identity by combining real data (e.g., a stolen SSN/Aadhaar number) with fake data (fabricated name, address, DOB) — or generating an entirely fictitious person. It is the fastest-growing and hardest-to-detect fraud type because there is no real victim to report the fraud.

How Synthetic Identities Are Created

graph TD
    A[Synthetic Identity Creation] --> B[Partial Synthetic<br/>Real SSN + fake name/DOB]
    A --> C[Full Synthetic<br/>All data fabricated]
    A --> D[AI-Generated<br/>Deepfake face + AI document + real/fake data]

    B --> E["Stolen SSN (child, deceased, immigrant)<br/>+ invented name + fake address"]
    C --> F["Generated SSN pattern + fake everything<br/>Passes format checks but no real person"]
    D --> G["StyleGAN face + forged ID document<br/>+ plausible but fabricated identity"]

    style D fill:#e53935,color:#fff

Lifecycle of Synthetic Identity Fraud

Phase What Happens Duration
1. Creation Build synthetic identity from combined real/fake data Days
2. Cultivation Open accounts, build credit history, make small legitimate transactions 6-24 months
3. Bust-out Max out all credit lines, take loans, cash out Days-weeks
4. Disappearance Abandon identity — no real person to collect from Permanent loss

Why It's Hard to Detect

Reason Details
No victim complaint Real identity theft triggers fraud alerts — synthetic doesn't
Passes format checks SSN/Aadhaar format is valid even if mismatched
Builds real history After cultivation phase, has legitimate-looking credit history
Face doesn't exist AI-generated face passes liveness but has no database match
Document may be real Real SSN on a forged document with fabricated name

Detection Methods

Method How It Helps
1:N face deduplication Same face across multiple identities
SSN/Aadhaar cross-reference Name-SSN mismatch (SSN belongs to different person)
Address analysis Multiple unrelated identities at same address
Credit history analysis Thin file, no history before a certain date
Device fingerprinting Same device used for multiple identities
Network analysis Connections between synthetic identities (shared addresses, phones, devices)
AI face detection Detect GAN-generated face artifacts

Key Takeaways

Summary

  • Synthetic identity fraud is the #1 emerging threat — $6B+ losses in US alone
  • No victim reports it — unlike stolen identity fraud, there's no alert
  • Cultivation phase (6-24 months) means losses are delayed and massive
  • Detection requires cross-referencing multiple signals: face dedup, database verification, device analysis, network analysis
  • AI-generated faces add a new dimension — synthetic people that have never existed