Use this framework to plan, execute, and document systematic adversarial testing that reveals AI vulnerabilities before deployment.

📋 1. Define Red Teaming Scope & Objectives

Establish clear boundaries and goals for your adversarial testing exercise.

Model & System Information

Model Details:

• Model Name/Version: _________________________
• Model Type: ☐ LLM ☐ Computer Vision ☐ Multimodal ☐ Other: __________
• Training Methodology: ☐ Supervised ☐ RLHF ☐ Constitutional AI ☐ Other: __________
• Deployment Stage: ☐ Pre-deployment ☐ Production ☐ Update/Iteration

Intended Use Cases:

• Primary use case: _________________________
• User groups: _________________________
• Deployment environment: _________________________

Red Teaming Objectives

Primary Testing Goals (select all that apply):

• ☐ Discover safety vulnerabilities and harmful output potential
• ☐ Test alignment boundaries and value misalignment scenarios
• ☐ Identify bias and fairness issues across demographic groups