> 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/online-fraud-detection/faq.md).

# FAQ

### **Product FAQs**

#### **1. What is VIDA Fraud Detection?**

VIDA Fraud Detection is an **AI-powered fraud prevention system** that detects **deepfake attacks, identity spoofing, and manipulated images** during the authentication process.

#### **2. How does VIDA detect deepfake attempts?**

VIDA leverages **advanced AI and ML algorithms** for **passive liveness detection** to analyze whether a person is physically present or using a spoofed artifact.

#### **3. What are Active and Passive Liveness Detection?**

VIDA supports both:

* **Passive Liveness Detection** – Runs in the background, analyzing selfies using AI/ML techniques without requiring user interaction. The system generates a **selfie score (0 to 1)** to determine authenticity.
* **Active Liveness Detection** – Requires users to perform simple actions (blinking, smiling, or head movement) to verify their presence and prevent fraud.

#### **4. How does VIDA’s Deepfake Defender API enhance fraud detection?**

VIDA’s **Deepfake Defender API** analyzes a selfie’s **liveness** in real-time and assigns a **liveness score**.\
The **selfie score** ranges from **0 to 1**, where:

* **0 (zero)** represents **maximum liveness**, indicating a highly confident live selfie.
* **1** represents **minimum liveness**, suggesting a potential spoof attempt.

The **success or failure** of the **liveness check** depends on the **liveness threshold score**. The default threshold is **0.95**, meaning any score **below 0.95** is considered a **successful liveness check**, while scores **above 0.95** may indicate a failed check due to possible spoofing or deepfake attempts.

#### **5. What is the VIDA Deepfake Shield SDK?**

VIDA’s **Deepfake Shield SDK** is an AI-powered **real-time face authentication and fraud prevention** solution. It protects against:

* **Printed photos and screen attacks**
* **2D masks and digital injection attacks**
* **Manipulated or spoofed images**

#### **6. What additional fraud prevention mechanisms does VIDA use?**

* **Image Manipulation Detection** – Identifies edited or tampered images.
* **Device-Based Security** – Ensures authentication happens only from trusted devices.
* **AI-powered anti-spoofing checks** – Detects **screen attacks, paper attacks, and deepfake attempts**.

#### **7. How does VIDA ensure secure data handling?**

VIDA performs:

* **Demographic Information Validation** – Matches user details (ID number, name, DOB) with official records.
* **Income Assessment** – Evaluates income data to assess financial credibility.
* **Regulatory Compliance** – Ensures adherence to security and privacy regulations.

#### **8. How does VIDA prevent fraud in high-risk industries?**

VIDA’s fraud detection solutions are ideal for:

* **Banking & Finance** – Prevents identity fraud in KYC processes.
* **E-commerce** – Blocks fake accounts and fraudulent transactions.
* **Telecommunications** – Detects SIM swap fraud and fake activations.
* **Government & Digital Identity** – Secures online identity verification

### **Integration FAQs**

#### **1. How can VIDA’s Deepfake Defender API be integrated?**

VIDA provides a **RESTful API** that allows businesses to:

* Send user selfie images for **liveness verification**.
* Receive a **liveness score** and **transaction ID** in response.

#### **2. What platforms support the VIDA Deepfake Shield SDK?**

The SDK is available for:

* **Android and iOS applications**
* **Web platforms**

#### **3. How do developers integrate the Deepfake Shield SDK?**

Developers can:

* Add **just a few lines of code** to enable deepfake detection.
* Customize UI elements and **detection parameters** (e.g., blinking, smiling, or head movement).

#### **4. What languages does the VIDA SDK support?**

VIDA SDK supports **English** and **Indonesian** for broader adoption.

####