AI Security Glossary
Plain-English definitions of the agentic-AI security and governance terms that matter — written for the people who have to secure agents, not just talk about them.
Agent security
- Agentic AI securityAgentic AI security is the practice of protecting autonomous AI agents across every surface they touch — prompt, retrieval, tool calls, session memory, agent-to-agent communication, and output — since each is a channel an attacker can exploit to hijack the agent or exfiltrate data.
- Prompt injectionPrompt injection is an attack in which adversarial text is fed to an AI model so the model follows the attacker's instructions instead of the developer's — by treating untrusted input as if it were a trusted command.
- RAG poisoningRAG poisoning is an attack that plants malicious content in a knowledge source so that when a retrieval-augmented generation (RAG) system retrieves it, the AI model treats the attacker's text as trusted context — a form of indirect prompt injection.