PromptGuard vs PyRIT

PromptGuard is a security layer between an application and its LLM provider, inspecting requests before they reach the model. It describes 15 detectors across six layers for risks such as prompt injection, jailbreaks, PII, data exfiltration, toxicity, fraud, secrets, malware and tool injection. The product says it detects and redacts 43 PII types using reversible tokenization. Its SDK can automatically instrument supported LLM calls with one initialization call while leaving existing provider code unchanged. Listed providers include OpenAI, Anthropic, Gemini, Azure, Bedrock, Mistral, Cohere, DeepSeek, Groq, HuggingFace, Ollama and vLLM. Deployment options include managed cloud, hybrid self-hosting and air-gapped installations. In zero-retention mode, prompt and response content is not stored; default security events include a shortened preview and content hash. PromptGuard says customer prompts, completions and documents are not used to train, fine-tune or evaluate models. The Free tier is permanent and includes 20,000 scans a month, one API key, one project and 24-hour log retention. Paid plans include a 14-day money-back guarantee, not a trial.

PyRIT is a free, open-source framework for automated and human-led red teaming of generative AI systems, aimed at assessing their security and safety. It supports single-turn and multi-turn attack strategies, including Crescendo, TAP, and Skeleton Key. Scenarios combine strategies and datasets for repeatable assessments of content harms, psychosocial risks, data leakage, and other objectives. Assessments can run through a command-line scanner, an interactive shell, the CoPyRIT graphical interface, or custom workflows. Documented targets include OpenAI, Azure, Anthropic, Google, Hugging Face, custom HTTP endpoints and WebSockets, and web apps tested with Playwright. Its modular components include targets, converters, scorers, memory, datasets, attacks, and scenarios. Converters can encode, obfuscate, translate, or semantically alter prompts, and transform among text, images, audio, video, and files. Scorers can return Boolean or normalized 0.0–1.0 scores using LLMs, Azure AI Content Safety, or custom logic. Built-in memory tracks conversations, scores, and attack results using SQLite or Azure SQL. PyRIT is self-hosted and requires a Python environment and configured AI endpoints.