Prompt Injection
| Term | prism:PromptInjection |
|---|---|
| Label | Prompt Injection |
| IRI | https://w3id.org/prism#PromptInjection |
| Type | rdfs:Class |
| Broader/Parent types | prism:AgenticRisk |
| Definition | Adversarial instructions embedded in content the agent processes (documents, web pages, tool outputs) redirecting it towards harmful actions, including data exfiltration through its tool access. |