Blog
AI DLP Research & Security Guides
FrameworkGenerative AI Governance: A Framework for Enterprise Security Teams
How to build a generative AI governance program that covers policy, technical enforcement, audit, and employee training — without blocking the productivity gains your business already depends on.
TechnicalAI DLP vs Network DLP: Why Traditional Data Loss Prevention Misses AI Prompts
Network DLP cannot intercept browser-submitted AI prompts without TLS inspection. This guide explains the architectural difference, why it matters, and what browser-native AI DLP does instead.
GuideHow to Stop Employees from Leaking Sensitive Data to AI Tools
A practical guide for security teams building controls around employee AI tool usage. Covers what data is at risk, how to enforce policy without killing productivity, and how to deploy without network changes.
GuideWhat is AI DLP? The Complete Guide to AI Data Loss Prevention
AI DLP (AI Data Loss Prevention) stops employees from sending sensitive data to ChatGPT, Claude, and Gemini. This guide explains how it works, why traditional DLP fails, and what to look for in an AI DLP solution.
CISOThe CISO's Guide to Generative AI Risk: What to Govern, What to Block, and What to Allow
A practical framework for CISOs building an enterprise generative AI governance program. Covers what data to block, which use cases to allow, and how to enforce policy without killing productivity.
GuideChatGPT Data Loss Prevention: How to Stop Sensitive Data from Reaching OpenAI
ChatGPT is the highest-volume AI data leakage surface in most organizations. This guide covers what data is at risk, why network DLP misses it, and how to configure effective ChatGPT DLP.
TemplateAI Acceptable Use Policy Template for Enterprise Teams
A ready-to-use AI acceptable use policy template covering approved tools, prohibited data types, enforcement, and employee responsibilities. Copy, customize, and deploy.
GuideAI Prompt Security: How to Keep Sensitive Data Out of AI Chat Interfaces
AI prompt security is the practice of preventing sensitive organizational data from being submitted to AI tools. This guide covers the threat model, detection approaches, and how to build a prompt security program.
EngineeringEngineering AI Security Starter: Stop Credentials and IP Leaking Through Your Dev Team
How to configure Pretzel to catch API keys, connection strings, and proprietary code before they reach AI coding assistants.
FintechFintech AI Risk Template: Keeping PCI, AML, and Trading Data Out of AI Tools
A practical Pretzel policy template for financial services teams. Covers credit card patterns, AML keywords, and MNPI detection.
LegalLegal AI Usage Policy: Protecting Attorney-Client Privilege in the Age of AI
How law firms and in-house legal teams can use AI tools without waiving privilege. Includes a practical Pretzel policy template.
HealthcareHIPAA AI Policy Template: What to Block Before Your Clinical Team Uses ChatGPT
A practical guide to configuring AI DLP for healthcare teams. Covers the 18 HIPAA PHI identifiers and how to enforce them with Pretzel.
Report5 Types of Data Your Team Is Accidentally Leaking to ChatGPT
We analysed 100,000 AI prompts. Here's what security teams found and how Pretzel stops it.