OWASP GenAI Security Project Gets Update, New Tools Matrix

OWASP published updated AI security guidance and solution maps for generative AI/LLMs and agentic AI, and released a GenAI Data Security list of 21 risks (including sensitive data leakage, data poisoning, exposure of agent identities/credentials, and unsanctioned data flows). The report highlights rapid change in the AI ecosystem, growth of agentic "swarms," insecure components (e.g., MCP servers), and calls for improved visibility, governance, and tailored defenses across development and deployment lifecycles.