Unverified50% confidenceFactExact time
The OWASP LLM Top 10 (2026) subdivides prompt injection into two categories: direct injection and indirect injection.
1
Sources
50%
Confidence
Long-term
Relevance
10/7/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedOWASP has listed prompt injection as the top security risk for large language model applications.61% similarVerifiedEMNLP is organized by SIGDAT (Special Interest Group on Linguistic Data and Corpus-based Approaches to NLP) under ACL57% similarUnverifiedDASA optimizes a set of continuous synthetic input embeddings and uses them directly for downstream fine-tuning, rather than using human-readable text.56% similarUnverifiedFew-shot在提示词中内嵌2-10个输入输出示例，可显著提升结构化输出的一致性55% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/983756API
curl https://kongchang.com/api/v1/knowledge/claims/983756MCP
get_claim(id=983756)