> For the complete documentation index, see [llms.txt](https://verogram.gitbook.io/whitepaper/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://verogram.gitbook.io/whitepaper/chapter-8-anti-fraud-and-risk-system.md).

# Chapter 8: Anti-Fraud & Risk System

### 8.1 Risk Scoring Model

Every user in Verogram has a dynamic risk score calculated from:

* Account age and activity history
* Task completion patterns
* Verification failure rate
* Behavioral anomalies (e.g., completing tasks too quickly)
* Connection patterns (e.g., multiple accounts linked to the same social profile)

### 8.2 Fraud Detection Mechanisms

* **Sybil Detection:** Identifying multiple accounts controlled by a single entity
* **Coordination Detection:** Detecting groups of users performing tasks in coordinated patterns
* **Bot Detection:** Identifying automated task completion through behavioral analysis
* **Rate Limiting:** Restricting the number of tasks a user can accept within a given timeframe

### 8.3 Consequences of Fraud Detection

* Reduced reward allocation
* Temporary or permanent account suspension
* Forfeiture of staked tokens (slashing)
* Permanent risk score elevation

### 8.4 Continuous Improvement

The anti-fraud system is designed to evolve continuously based on:

* New attack patterns observed
* Data analysis of completed campaigns
* Feedback from advertisers regarding engagement quality