All articles
-
How to Scan an LLM for Prompt Injection: Tools, Method, and Limits
A working guide to scanning LLM applications for prompt injection: offline probe suites like garak and PyRIT, runtime classifiers like Azure Prompt Shields, and what a clean scan does and does not prove.
-
Best Tools to Test AI Chatbot Security in 2026
Garak, PyRIT, Promptfoo, Giskard, and Lakera Red compared as tools to test AI chatbot security across full conversations, not single-shot prompts.
-
How to Detect Prompt Injection Vulnerabilities in LLM Apps
A practitioner's guide to detecting prompt injection vulnerabilities: input classifiers, output validators, structural controls, and red-team testing.
-
OWASP LLM Top 10 Scanning Checklist: What to Test at Each Layer
An OWASP LLM Top 10 scanning checklist: what to test for each 2025 risk, which open-source scanners cover it (garak, PyRIT, promptfoo), and the gaps.
-
Best LLM Vulnerability Scanners 2026: Garak, PyRIT, Promptfoo
A practitioner's guide to the best LLM vulnerability scanners in 2026: Garak, PyRIT, Promptfoo, and Mindgard, plus the specialist and runtime layers around them.
-
Open Source LLM Red Teaming Tools: PyRIT, Garak, and HarmBench
A practitioner's guide to the main open source LLM red teaming tools: what PyRIT, Garak, HarmBench, and TextAttack cover and where they fall short.
-
LLM Red Teaming Tools: garak vs PyRIT vs Promptfoo in CI
Compare garak, PyRIT, and Promptfoo as LLM red teaming tools: what each scans, how each fits a CI gate, and which to run per commit vs nightly.
-
Choosing an LLM Guardrail: Llama Guard, NeMo, Guardrails AI
A decision guide for picking an LLM guardrail in 2026 — Meta's Llama Guard 4, NVIDIA's NeMo Guardrails, and Guardrails AI.
-
Rebuff, Vigil, LLM Guard: Open Source Prompt Injection Tools
Rebuff is archived, Vigil is alpha, LLM Guard is maintained. Compare detection layers, licenses and status, and see which one to run in production.
-
LLM Guard: Input and Output Scanning for Production LLM Apps
A practical breakdown of LLM Guard by Protect AI: its input and output scanners, how the scan pipeline works, and where it fits as a runtime control.
-
PyRIT: Microsoft's AI Red-Teaming Framework, Explained
A technical breakdown of PyRIT, Microsoft's red-teaming framework for generative AI, covering targets, datasets, orchestrators, converters, and scorers.
-
False Positive Cost in Refusal Systems: Measure and Tune
Practical methods for quantifying refusal false positives in LLM products: eval design, baseline rates, threshold tuning, and the regression suite.
-
Garak LLM Vulnerability Scanner: How It Works and When to Use It
A technical breakdown of the garak LLM vulnerability scanner: its probe architecture, attack categories, CLI workflow, and how it fits a pre-deployment gate.
-
Classifier-on-Output: Catching Misbehavior Post-Generation
How production teams use post-generation classifiers to catch what input filters and refusal training miss, with the architectures and their real costs.
-
Llama Guard vs NeMo vs OpenAI Moderation: Production Tradeoffs
A comparison of Llama Guard, NeMo Guardrails, and the OpenAI Moderation API on policy fit, latency, cost, failure modes, and when to pick each one.