Why Your LLM Provider's Built-In Guardrails Are Not Enough
OpenAI, Anthropic, Google, and AWS offer PII filtering, but it all runs after your data leaves your network. Why filtering out is not the same as never sending.
Redact Both Sides of the LLM: Inputs and Outputs
Prompt redaction protects the provider boundary. Output scanning protects everything downstream. An LLM needs PII redaction on both sides, not one.
Can an LLM Leak Its Training Data, and Why You Cannot Un-Train PII
Research shows LLMs memorize and leak training data, and unlearning it afterward is unreliable. The dependable control is to redact before training.
How to Redact PII Before Sending to an LLM: Chat, RAG, and AI Agents
Redact PII and PHI before any prompt reaches an LLM, keeping data on your own infrastructure. A self-hosted pipeline for chat, RAG, and AI agents.
Prompt Engineering for Privacy: Practical Patterns for Not Leaking PII
Every prompt sent to an LLM is a data egress point. Six concrete patterns for structuring prompts, redacting inputs, and scanning outputs so PII doesn't leak.
Using an LLM or Pattern-based Rules for PII/PHI Redaction
Comparing the two main approaches to redacting PII and PHI: an LLM versus pattern-based rules. How each handles accuracy, cost, and GDPR or HIPAA compliance.
Why Using an LLM to Redact PII and PHI is a Bad Idea
Lots of posts show how to redact PII and PHI text with a large language model (LLM). Can we really just let an LLM handle it? Here is why that is a bad idea.
Philter as an AI Policy Layer
An AI policy layer inspects AI-generated text to prevent sensitive information from being exposed, removing names, addresses, telephone numbers, and more.
Medical Chatbot — User Input Redaction
LLM Training Data Preparation