How to Craft AI-Generated Attack Vector

What is VectorAI™?

VectorAI is an add-on module that crafts AI-generated DDoS attack vectors, which are then delivered by RADAR to validate how enterprise defenses perform against advanced attack scenarios.

Unlike traditional attack libraries that rely on predefined attack patterns, VectorAI generates AI-assisted attack-vector variations based on the selected Profile, Attack Layer, Port, and AI Model. This enables organizations to expand validation coverage beyond static attack libraries and test more adaptive attack scenarios.

The selected Profile provides VectorAI with the relevant protection context used to generate more targeted attack-vector recommendations. This helps organizations evaluate how their protections perform against attack scenarios that are tailored to the environment being validated.

VectorAI extends the capabilities of RADAR by generating additional attack-vector variations that can be reviewed, managed, and used in simulations alongside RADAR’s predefined attack library. This helps organizations broaden validation coverage and maintain confidence in their protection readiness as attack techniques continue to evolve.

What You'll Learn on This Page

On this page, you will learn how to craft AI-generated attack vectors using VectorAI, understand the required generation parameters, and incorporate generated attack vectors into simulation workflows for validation and testing.

Craft AI-Generated Attack Vector

Follow these steps to craft a new AI-generated attack vector in the RADAR platform:

1. In the left-hand menu, navigate to VectorAI and select Attack Generator.
2. In the Personalize Your Attack panel, configure the Scope fields:
   
   • In the Profile dropdown, select the profile that represents the relevant protection and validation context for the generated attack vector.
   • In the Port field, enter the port to be used as part of the attack-vector generation.
     
   • In the Attack Layer dropdown, select the attack layer for the generation request.
     VectorAI currently supports Layer 7 – Application attack-vector generation.
3. In the top-right corner of the Generate DDoS Attack panel, select an AI model from the model selector. The default model is Sonnet 4.6 Extended.
4. Click Generate.

The Generate DDoS Attack panel displays the generation progress. Generation may take up to 10 minutes to complete. You can click Cancel at any time to stop the process.

Understanding the Generation Fields

The Personalize Your Attack panel contains the fields that define the scope and context of the attack vector VectorAI will craft. These fields directly influence the attack-vector recommendations generated by the selected AI model.

Profile
The mitigation profile that defines the protection context for the generation. VectorAI uses the selected profile to craft attack-vector recommendations that are tailored to the environment being validated.

Port
The target port number for the generated attack vector.

Attack Layer
The OSI layer the generated attack vector will target.

AI Model
The AI model used to craft the attack vector. The model selector is located in the top-right corner of the Generate DDoS Attack panel. The default model is Claude Sonnet 4.6. The selected AI model affects generation only and does not affect simulation execution or validation results.

• Claude Sonnet 4.6 – Optimized for faster generation and balanced attack-vector recommendations, making it suitable for most use cases
• Claude Opus 4.7 – Designed for more advanced reasoning and deeper analysis, which may produce more detailed attack vectos
• Claude Haiku 4.5 - Fastest, most lightweight model, built for quick, low-latency generation while maintaining strong output quality
• OpenAI GPT 5.5 - A general-purpose model, typically positioned for broad reasoning and instruction-following across varied tasks
• Google Gemini 3.1 Pro - Google's mid-to-high tier Gemini model, generally positioned for strong reasoning with multimodal capabilities
• xAI Grok 4.3 - Typically positioned around fast, direct responses and real-time knowledge integration
• DeepSeek V4 Pro - Advanced model tier, generally positioned for strong reasoning performance at competitive efficiency

All models consume one credit per generation request.

What Happens After Generation?

After an attack vector is generated successfully, it appears in the Generate DDoS Attack panel with the following details:

• Attack Name – The name assigned to the generated attack vector
• Generation Date and Time – The date and time the attack vector was created
• Attack Summary – A high-level overview of the attack behavior and validation objective
• Attack Layer – The attack layer selected during generation
• Attack Description – Detailed information about the attack scenario, including the conditions and protection gaps it is designed to validate

The generation attempt is recorded in the History panel, which tracks all generation attempts including failed ones. Successfully generated attack vectors are automatically added to the Attack Library, where you can review them.

Generating an attack vector does not validate protection by itself. Protection is validated only after the attack vector is included in a scheduled simulation and executed against the configured Targets. Simulation results can then be used to assess how the environment responds to the generated attack scenario.

On the upper-right side of the page, the following actions are available:

"New" Button
Click New to start a new attack generation workflow. This returns you to the attack generation screen, where you can create an additional AI-generated attack vector.

"History" Button
Click the History icon to view all attack generation attempts. The History window displays:

• Successfully generated attack vectors
• Failed generation attempts
• Generation status and details

You can also use the Search field to locate generated attack vectors by name.

Schedule a Simulation Using an AI-Generated Attack Vector

To validate the generated attack vector, navigate to the Scheduler page and create a simulation that includes the new attack vector.

Follow these steps to schedule a simulation using an AI-generated attack vector:

1. In the left-hand menu, navigate to Scheduler.
2. Click Add Event.
3. Under the Coverage tab:
   1. In the Specific Attack Vectors field, select the AI-generated attack vectors you want to include in the simulation.
   2. Alternatively, enable Include VectorAI-Generated Attack Vectors to automatically include all available AI-generated attack vectors that are compatible with the selected simulation configuration.
4. Click Next to continue configuring the remaining simulation settings.
5. Complete the event configuration, then click Save to create the simulation.

Once scheduled, the simulation will execute using the selected AI-generated attack vectors and the configured Targets.

What's Next

Now that you’ve crafted AI-generated attack vector, proceed to the How to View the AI-Generated Attack Library to review and simulate all generated attack vectors.