> For the complete documentation index, see [llms.txt](https://docs.vida.id/identity-stack/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.vida.id/identity-stack/identity-operation-center/fraud-ops/liveness-fraud-ops.md).

# Liveness Fraud Ops

## What is Liveness Fraud Ops?

VIDA Liveness Fraud Ops is a **near-real-time manual fraud review** system for VIDA liveness service. When your system accepts a user's liveness check based on VIDA’s automated evaluation, Fraud Ops **intercepts that session for deeper inspection,** especially when the session score passes the AI threshold but may still exhibit signs of spoofing.

With a specially-trained & quality-controlled human agents, VIDA does secondary review for transaction passing AI-based liveness checking, flags whether the transaction is genuine or fraudulent, and sends you the result via webhook or API, all within up to 15 minutes. If the session is flagged as fraud (e.g., deepfake), your backend can immediately take action according to your business process (e.g.: block the transaction, blacklist the user, or so on)

### Key Benefits

<table><thead><tr><th width="244.9375">Benefit</th><th>Key Value</th></tr></thead><tbody><tr><td><strong>Rapid Fraud Interception</strong></td><td>Detect and act on suspicious liveness sessions within 15 minutes.</td></tr><tr><td><strong>Additional Human Expertise for AI edge cases</strong></td><td>Specially-trained &#x26; quality-controlled agents do secondary checking to identify AI biases &#x26; bypasses </td></tr><tr><td><strong>Protection Against High-Impact Fraud</strong></td><td>Especially valuable for enterprises in finance, fintech, telcos, and e-commerce dealing with high-value transactions.</td></tr><tr><td><strong>Clear Spoof Reason</strong></td><td>You will receive detailed fraud classification and justification (e.g.: Injection Attack - Watermark)</td></tr></tbody></table>

## How Fraud Ops Works

* **Trigger**: A liveness session that passes VIDA’s AI (accepted) check is routed to a **human reviewer** if:
  * Fraud Ops is enabled for your integration configuration.
  * The liveness score is within accepted threshold (e.g., <0.95).
* **Agent Review**: A trained VIDA agent reviews the selfie:
  * Is it a real human, or a spoof (deepfake, static image, mask, etc.)?
* **Result Notification**:
  * If photo turns out to be spoof → a webhook notification is sent to your backend.
  * You can also use our Livenss Status API to query the result.
* **Your Action**: Once a spoof notification is received, you can take action depending on your business process (e.g.: block the transaction, blacklist the user, or so on)