Unverified50% confidenceFactExact time
XPIA (Cross-domain Prompt Injection Attack) involves an attacker pre-embedding hidden instructions in documents that will be retrieved, so that when the documents are injected into the model as context, the malicious instructions may be mistakenly executed as legitimate ones.
1
Sources
50%
Confidence
Long-term
Relevance
10/11/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedIndirect injection refers to attackers hiding malicious instructions in external content that the model will read, such as webpage body text, email body text, or document content.70% similarVerifiedPrompt injection attacks induce AI agents to execute malicious operations by embedding text disguised as system instructions in code comments, documentation, web content, or even file names.69% similarUnverifiedStored Cross-Site Scripting (Stored XSS) is persistent and propagating — attackers inject malicious scripts into a database or file system, and a single injection can affect all subsequent visitors.67% similarVerifiedIndirect prompt injection attacks plant hidden instructions within normal business data to induce Agents into executing unauthorized operations, representing a security risk at the semantic level rather than the code level.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/998146API
curl https://kongchang.com/api/v1/knowledge/claims/998146MCP
get_claim(id=998146)