OWASP LLM09 (2023) Overreliance

What is OWASP LLM09 (2023) Overreliance?

Overreliance captured the human factor: teams treating fluent answers as facts in high-stakes workflows without checks, escalation, or source verification.

What this looks like in production

Part of this risk is reframed in LLM09 (2025) Misinformation, which emphasizes the security impact of confident wrong outputs, not only user behavior.

Further reading on Giskard

Official OWASP reference

Get AI security insights in your inbox