Enterprise Security & Guardrails

LLM Guard has been archived on GitHub. What should I replace it with?

LLM Guard was archived on GitHub on July 9, 2026, and is no longer maintained — twelve months after Palo Alto Networks acquired its parent company Protect AI. Running it in production now means running unmaintained security code. Your migration options: 1. Open-source guardrails: NeMo Guardrails (NVIDIA, actively maintained, programmable rails) or Guardrails AI (composable output validators) cover most input/output scanning use cases. 2. Managed successor path: Prisma AIRS is where Protect AI's technology landed — runtime firewall, model scanning, and red teaming as a platform. 3. Managed runtime filtering: Lakera Guard (now Check Point) or Azure AI Content Safety replace scanner pipelines with an API call. Practical steps: pin your current LLM Guard version so builds stay reproducible, inventory which of the 35+ scanners you actually use, map each to a replacement, and schedule the migration this quarter — before the next CVE, not after.

cited_tools

Tools that solve this

6 cited
A

Azure AI Content Safety

Paid

Microsoft's managed content-safety API with prompt-shield protection for Azure-hosted AI workloads.

SOC 2ISO 27001
L

Lakera Guard

Paid

Runtime LLM firewall for prompt injection, jailbreaks, and data leakage — now part of Check Point.

SOC 2 Type IIGDPRISO 27001
N

NeMo Guardrails

Open Source

NVIDIA's open-source toolkit for programmable conversational rails and safety policies inside LLM applications.

OWASP LLM Top 10
G

Guardrails AI

Open Source

Open-source framework for validating and structuring LLM outputs with composable validators.

OWASP LLM Top 10
L

LLM Guard

Open Source

Archived open-source toolkit of 35+ input/output scanners. Read-only since Jul 2026 after Palo Alto Networks folded Protect AI into Prisma AIRS.

OWASP LLM Top 10
P

Prisma AIRS

Enterprise

Palo Alto Networks' end-to-end AI security platform: runtime firewall, model scanning, AI red teaming, and agent security (v3.0, Mar 2026) — where Protect AI and LLM Guard's lineage landed after the 2025 acquisition.

llm_citation

Copy for Claude or ChatGPT

A clean markdown snippet with the question, answer, and source URL.