# What is Guardrails AI? Definition and meaning

A framework that checks model responses against validators for structure, safety and content rules.

## Definition

**Guardrails AI**: A framework that checks model responses against validators for structure, safety and content rules.

In short: Validation of language model inputs and outputs.

## When it is used

Enforcing output formats and safety constraints in applications.

## Where it fits

Guardrails AI belongs to **LLM safety** in the [Generative AI and LLMs](/topics/generative-ai-llms) topic. These are the other terms in the same group:

- **[NeMo Guardrails](/glossary/nemo-guardrails)**: A toolkit that adds topic, safety and security rails to LLM applications using a dedicated rule language.
- **[Presidio](/glossary/presidio)**: Microsoft's open-source toolkit for finding and anonymizing personally identifiable information in text and images.
- **[Red-teaming](/glossary/red-teaming)**: Systematically testing a model or application with adversarial inputs to find unsafe or unintended behaviour.

## Learn more

The full guide to **[Guardrails AI](/tools/guardrails-ai)** explains how it works, when to use it and how it compares with the alternatives, with examples. The [glossary](/glossary) lists every term on the site from A to Z.

Previous term: [Guardrails](/glossary/guardrails) · Next term: [Hadoop HDFS](/glossary/hadoop-hdfs)
