OWASP Top 10 LLM Risks 2025: Key AI Security Updates - Qualys
The OWASP Top 10 for LLM Applications 2025 introduces critical updates, including new entries like System Prompt Leakage, which exploits the exposure of sensitive data embedded in application prompts. It also highlights Vector and Embedding Weaknesses, focusing on vulnerabilities within Retrieval-Augmented Generation (RAG) and other embedding-based LLM architectures.
Source: Original Report ↗