OWASP Top 10 for Large Language Model Applications

Our summary

A community-built catalog of the most critical security risks for LLM applications — prompt injection, insecure output handling, training-data poisoning, and more — with mitigations for each.

Why it matters

The de-facto checklist teams reach for when threat-modeling an LLM deployment.

Cited by these methods

Related findings (1)

• Indirect prompt injection via retrieved contentCritical

  Instructions hidden in documents, web pages or tool outputs can override the system prompt when ingested by the model.

Cite this