[KongchangAI]
Unverified50% confidenceFactExact time

2024年安全研究员Johann Rehberger演示了通过在公开网页中嵌入对人类不可见但对AI可读的指令文本成功劫持多个商业RAG系统输出的间接提示注入攻击

1
Sources
50%
Confidence
Long-term
Relevance
8/7/2026
First Seen

Sources

Related Claims

Cite This Claim

Stable URI
https://kongchang.com/claim/700787
API
curl https://kongchang.com/api/v1/knowledge/claims/700787
MCP
get_claim(id=700787)