NVIDIA Safety for Agentic AI

NVIDIA AI Blueprint: Safety for Agentic AI

Deprecation: This blueprint is no longer actively maintained as of April 22, 2026. We recommend using NeMo Microservices instead: NeMo Auditor for pre-deployment vulnerability scanning, NeMo Guardrails for runtime safety protection, and Safe Synthesizer for safe training data generation.

About Safety for Agentic AI

As large language models (LLMs) increasingly enable agentic AI systems capable of autonomous reasoning and tool use, they also introduce critical safety risks, including goal misalignment, hallucinations, and prompt injections. Enterprises are challenged to harness open-weight models' flexibility without compromising on trust, security, or compliance. As regulations tighten across regions and industries, non-compliance becomes a persistent challenge.

With this safety recipe, enterprises can now confidently adopt open models, aligned to their policy. Start with model evaluation using garak vulnerability scanning with curated risk prompts, benchmarking against enterprise thresholds. Then, post-train using recipes and safety datasets to close critical safety and security gaps. Deploy the hardened model as a trusted NVIDIA NIM and then add inference run-time safety protection with [NVIDIA NeMo Guardrails](https://developer.nvidia.com/nemo-guardrails/) that actively block unsafe model behavior. With continuous monitoring, and collaboration between AI and risk teams, model safety becomes enforceable, not aspirational.

For guidelines and suggestions about the process, refer to [Best Practices for Developing a Model Behavior Guide](https://github.com/NVIDIA-AI-Blueprints/safety-for-agentic-ai/blob/main/docs/best-practices-model-behavior-guide.md).

This repository is what powers the [build experience](https://build.nvidia.com/nvidia/safety-for-agentic-ai), helping you harden security at every stage of the AI development lifecycle.

...