πŸ“– Step 9: AI/LLM#342 / 350

Red Teaming

Red Teaming

πŸ“–One-line summary

Testing that deliberately probes a model's vulnerabilities to verify its safety.

πŸ’‘Easy explanation

A team that plays hacker, deliberately attacking an AI to find its weaknesses. They simulate worst-case scenarios before launch to strengthen defenses.

✨Example

μ˜λ„μ μœΌλ‘œ 약점 μ°ΎκΈ°

πŸ”΄ λ ˆλ“œ νŒ€ (곡격)

β†’ "μœ ν•΄ μ½˜ν…μΈ  μœ λ„ μ‹œλ„"

β†’ "κ°œμΈμ •λ³΄ μΆ”μΆœ μ‹œλ„"

🟒 κ²°κ³Ό: 취약점 3건 발견 β†’ 패치 μ™„λ£Œ