Shadow AI Detection & Regularization
Your AI system inventory only covers what you already know exists. This methodology finds the AI your workforce is already using that nobody has declared.
What's included
- PDF with the 3-route detection methodology: employee survey, SaaS subscription expense review, and what to look for if you already have network monitoring tools (CASB, DLP, proxy)
- Employee survey template, ready to adapt and send
- Excel with a findings log and regularization plan (approve, replace with an approved alternative, or ban)
- Legal Notice
Why this document exists
Almost every company doing its first AI system inventory discovers the same thing: the official list of approved systems is much shorter than the AI the workforce is actually using. Someone in marketing has been using generative AI to draft content for months, an analytics team connected an AI API to customer data without going through security, an entire department pays for an AI tool subscription that was never assessed. It's not bad faith — the official route either didn't exist or was slower than solving the problem on their own. This document doesn't punish that reality, it finds it and regularizes it.
Frequently asked questions
What exactly is Shadow AI?
AI tools that employees or teams use in their daily work without going through any approval process, risk review, or formal registration. It's not malice — it solves a real problem faster than waiting for a formal process.
How is this different from the AI Systems Register?
The register assumes you already know which AI systems exist and just need to document them. This product is the step before: how to find AI usage nobody has declared yet.
Do I need technical network monitoring tools?
Not essential — the methodology includes non-technical routes (survey, expense review, interviews), plus suggestions on what to look for if you already have CASB, DLP, or proxies.
What format is it delivered in?
PDF with the 3-route detection methodology, an employee survey template, and Excel with a findings log and regularization plan.