AI glossary · Ethics, safety and privacy
Guardrails
Guardrails are checks and limits built into an AI product to stop harmful, off-topic or unauthorised outputs and actions.
Also called AI guardrails, content filters, safety filters
In more detail
They can be instructions in a system prompt, filters that scan inputs and outputs, rules about which tools an agent may use, or approval steps before risky actions.
No guardrail is perfect; people find ways around them with jailbreaks and prompt injection, so they work best in layers.
Related terms
AI jailbreakAn AI jailbreak is a prompt designed to trick a chatbot into ignoring its safety rules and producing content it is meant to refuse.Prompt injectionPrompt injection is an attack where hidden instructions in content an AI reads, such as a web page or email, trick it into doing something it shouldn’t.AI safetyAI safety is the work of making sure AI systems behave reliably and don’t cause harm, from everyday mistakes to serious misuse and long-term risks.Human in the loop (HITL)Human in the loop means a person reviews, approves or corrects an AI system’s work at key points, instead of letting it act entirely on its own.
Browse related tools
AI tools listed on BroUseAI in these categories.
