# System Prompt Leakage: definition and example

URL: https://airuntimesecuritytools.com/glossary/system-prompt-leakage/
Updated: 2026-11-16

System prompt leakage is the exposure of the hidden instructions that steer a model, which can reveal rules, roles and any secrets placed in them. OWASP lists it among its top risks for LLM applications. The lasting fix is to keep secrets and authorization out of the prompt, since filtering cannot guarantee the text stays hidden.

## Example

OWASP says a system prompt should not be considered a secret, nor used as a security control.

## Sources

- [OWASP LLM07:2025 System Prompt Leakage](https://genai.owasp.org/llmrisk/llm072025-system-prompt-leakage/) (genai.owasp.org), read 2026-11-16

## Related pages

- [Sensitive Information Leakage](/glossary/sensitive-information-leakage/)
- [Output Filtering](/glossary/output-filtering/)
- [Canary Token](/glossary/canary-token/)
- [All glossary terms](/glossary/)
- [SOAR Playbook](/glossary/soar-playbook/)
- [Telemetry](/glossary/telemetry/)
- [Guides](/guides/)
