LLMMedium
Bing Chat's Initial Prompts Revealed by Early Testers Through Prompt Injection
- Occurred
- Feb 8, 2023
- Published
- Feb 10, 2023
- Source
- AI Incident Database
Summary
Early testers of Bing Chat successfully used prompt injection to reveal its built-in initial instructions, which contains a list of statements governing ChatGPT's interaction with users.
Developers
- Microsoft
- OpenAI
Deployers
- Microsoft
Harmed parties
- Microsoft
Classification
Category confidence: 90%
Severity confidence: 85%
Rationale: Prompt injection attack on an LLM chatbot revealing system instructions without causing physical harm or major financial loss.
Classifier: Qwen/Qwen3.5-35B-A3B-FP8