OWASP Top 10 LLM Risks Explained

### Executive summary: The OWASP Top 10 for LLM Applications (2025) outlines the leading security risks introduced by production use of large language models — including prompt injection, sensitive data leakage, supply-chain and data/model poisoning, improper output handling, excessive agency, system prompt leakage, vector/embedding weaknesses, misinformation, and unbounded resource consumption — and recommends familiar security controls (least privilege, input validation, supply-chain vetting, monitoring, and human oversight) alongside AI-specific testing to reduce exposure.