Red Teaming
Full Form: AI Red Teaming
Category: AI Safety
📖 Definition
Red teaming tests AI systems by attempting to make them fail, behave badly, or reveal harmful outputs. It helps identify vulnerabilities before deployment.
🔑 Key Points
- Deliberate attempts to find AI failure modes
- Tests for harmful outputs, biases, and vulnerabilities
- Used by AI labs before releasing models
- Can be internal or external (adversarial testing)
💡 Why It Matters
Red teaming helps identify and fix AI safety issues before they cause harm. It's a critical part of responsible AI development.