Skip to content
CyberSimStart training

Theory · 9 min

The OWASP Top 10 for LLM Applications 2025

Name the ten LLM risks and recognise each in an incident description.

OWASP’s 2025 list: LLM01 Prompt Injection, LLM02 Sensitive Information Disclosure, LLM03 Supply Chain, LLM04 Data and Model Poisoning, LLM05 Improper Output Handling, LLM06 Excessive Agency, LLM07 System Prompt Leakage, LLM08 Vector and Embedding Weaknesses, LLM09 Misinformation, LLM10 Unbounded Consumption.

Improper Output Handling (LLM05) is the classic web bug in a new place: model output passed unchecked into a browser, a shell or a database query becomes cross-site scripting, command injection or SQL injection.

Vector and Embedding Weaknesses (LLM08) are about retrieval-augmented generation (RAG): poisoned or over-shared documents in the knowledge base the model searches.

Unbounded Consumption (LLM10) covers runaway cost and denial of service: a public endpoint anyone can make generate expensive answers all day.

Your turn: match observations. Which OWASP LLM risk is each incident?

Do the interactive check Free, in the browser.

Sources

Part of