Cybersecurity glossary
What is prompt injection?
Also known as: indirect prompt injection, jailbreak
Definition
Prompt injection is an attack on AI applications in which crafted input, typed directly or hidden in content the model reads, overrides the developer’s instructions.
Why it matters
It is ranked first in the OWASP Top 10 for LLM Applications. Mitigations focus on limiting impact: least-privilege tool access, treating output as untrusted, separating instructions from data, and human approval for sensitive actions.
Common questions about prompt injection
Sources
Related Guides
GuideAI Security
How to Conduct LLM Security Testing for Web Applications
Learn how to perform LLM security testing to identify prompt injection, sensitive data leakage, and insecure output handling before production deployment.
Read more GuideAI Security
How to Deploy a Secure AI Chatbot for Small Business
Learn how to configure, test, and run a secure AI chatbot for your small business. Protect customer data, prevent prompt injection, and limit model access.
Read more GuideWeb Security
A Practical Guide to HTTP Security Headers for Web Developers
Why modern web applications need CSP, HSTS, and X-Content-Type-Options, and how to configure them properly in Next.js and Nginx.
Read more GuideWeb Security
How to Protect Your Projects From Malicious npm Packages (and What GitHub's New Package Firewall Does)
GitHub launched a package firewall that blocks risky npm packages before they install. Learn how it works, who can use it and 10 steps any team can take today.
Read more