LLM Guard by Protect AI is a comprehensive tool designed to fortify the security of Large Language Models (LLMs).
Documentation | Playground | Changelog
By offering sanitization, detection of harmful language, prevention of data leakage, and resistance against prompt injection attacks, LLM-Guard ensures that your interactions with LLMs remain safe and secure.
LLM Guard helps teams protect AI systems by enforcing security controls at the most vulnerable points of LLM usage:
- Before model inference (input guardrails)
- After model inference (output guardrails)
- During execution and tool usage (runtime guardrails)
It is designed for production AI systems, not just experiments.
Traditional application security tools are not designed for AI systems.
LLMs:
- Accept untrusted natural language input
- Produce non-deterministic outputs
- Can leak sensitive data through responses
- Can be manipulated to bypass policies at runtime
Without guardrails, LLM-powered systems introduce new attack surfaces that conventional security controls cannot detect or block.
LLM Guard is built around real-world LLM security threats, including:
- Prompt Injection (direct and indirect)
- Jailbreak attacks and role manipulation
- Sensitive data leakage in responses
- Training data exposure
- Unauthorized tool or function execution
- Policy and compliance violations at inference time
These risks exist even when models are hosted by trusted providers.
LLM Guard enforces security using layered guardrails:
-
Input Guardrails
Validate, sanitize, and assess prompts before they reach the model. -
Output Guardrails
Inspect model responses for sensitive data, policy violations, or unsafe content. -
Runtime Guardrails
Enforce security policies during execution, including tool calls and agent workflows.
| Control Type | AI Layer | Threats Mitigated |
|---|---|---|
| Input Validation | Prompt | Prompt injection, jailbreaks |
| Output Filtering | Response | Data leakage, unsafe outputs |
| Runtime Policies | Execution | Unauthorized actions, agent misuse |
| Governance Rules | System | Compliance and policy violations |
LLM Guard can be used to secure:
- Enterprise chatbots
- RAG (Retrieval-Augmented Generation) pipelines
- AI agents and autonomous workflows
- Internal AI assistants
- Customer-facing generative AI applications
Begin your journey with LLM Guard by downloading the package:
pip install llm-guardImportant Notes:
- LLM Guard is designed for easy integration and deployment in production environments. While it's ready to use out-of-the-box, please be informed that we're constantly improving and updating the repository.
- Base functionality requires a limited number of libraries. As you explore more advanced features, necessary libraries will be automatically installed.
- Ensure you're using Python version 3.9 or higher. Confirm with:
python --version. - Library installation issues? Consider upgrading pip:
python -m pip install --upgrade pip.
Examples:
- Get started with ChatGPT and LLM Guard.
- Deploy LLM Guard as API
- Anonymize
- BanCode
- BanCompetitors
- BanSubstrings
- BanTopics
- Code
- Gibberish
- InvisibleText
- Language
- PromptInjection
- Regex
- Secrets
- Sentiment
- TokenLimit
- Toxicity
- BanCode
- BanCompetitors
- BanSubstrings
- BanTopics
- Bias
- Code
- Deanonymize
- JSON
- Language
- LanguageSame
- MaliciousURLs
- NoRefusal
- ReadingTime
- FactualConsistency
- Gibberish
- Regex
- Relevance
- Sensitive
- Sentiment
- Toxicity
- URLReachability
LLM Guard is an open source solution. We are committed to a transparent development process and highly appreciate any contributions. Whether you are helping us fix bugs, propose new features, improve our documentation or spread the word, we would love to have you as part of our community.
- Give us a ⭐️ github star ⭐️ on the top of this page to support what we're doing, it means a lot for open source projects!
- Read our docs for more info about how to use and customize LLM Guard, and for step-by-step tutorials.
- Post a Github Issue to submit a bug report, feature request, or suggest an improvement.
- To contribute to the package, check out our contribution guidelines, and open a PR.
Join our Slack to give us feedback, connect with the maintainers and fellow users, ask questions, get help for package usage or contributions, or engage in discussions about LLM security!
We're eager to provide personalized assistance when deploying your LLM Guard to a production environment.