AIS

Prompt injection

Direct prompt injection is an attacker typing malicious instructions straight into an AI chat; indirect prompt injection hides those instructions in a document, webpage, or email the AI is asked to read, so it executes them without the user ever seeing them.

For small firms

Any AI tool that reads outside content — email, the web, uploaded files — needs this on the risk register, not just the chat box itself.

Related terms

AIS Jailbreak Crafting input that talks an AI model into ignoring its own safety rules. AIS Guardrails Technical controls that keep an AI system's inputs and outputs within bounds. AIS OWASP Top 10 for LLM Applications A maintained list of the most significant security risks in LLM applications.
← Back to AI security