How To Run AI Inventory Workshop

Most organisations are using more AI than they realize. Some AI is built into the product. Some is buried inside vendor tools. Some is being used informally by teams for support, sales, engineering, marketing, or operations.

Before you can assess AI Act exposure, privacy risk, vendor risk, or internal governance gaps, you need one thing first: a clear AI inventory. The fastest way to build an inventory is a workshop with the right people in one room.

Who should be there

Bring the people who know where AI is actually being used:

• Head of Engineering or CTO

• Head of Product

• Legal or compliance representative

• Security lead

• Data Privacy Lead

• Operations or IT representative

• Customer-facing team representative (sales, marketing and/or customer success)

Before The Workshop

Prepare a simple spreadsheet using our AI Inventory Template.

The rule for the session is simple: when in doubt, add it.

The Workshop Structure

1. Align On What Counts as AI

Start with a broad definition and define what counts as AI. This should include but not limited to Machine Learning (ML), generative AI, predictive models, recommendation systems, scoring tools, chatbots, co-pilots, AI assistants, automated decision-making and statistical models. Walk through a few examples to align understanding.

2. Map Internal AI Use

Identify every place employees use AI internally. Look across engineering, product, sales, customer success, marketing, security, operations, finance, legal, and HR. Capture every system, tool, and use case where the company has built or deployed AI internally.

3. Review Vendor-embedded AI

Walk through the vendor list. Flag every vendor with AI features (Copilot, AI Assistant, Smart Suggestions, Predictive Insights, Auto Summary, etc.). For each one, confirm whether the feature is enables, who uses it, what data it touches, and whether vendor documentation is available.

4. Map Product AI and Triage

Review every AI-enabled feature in your product. Capture risk-relevant detail. Wrap with a triage discussion of which systems need the deepest follow-up.

5. What You Should Have By The End

A good AI Inventory Workshop should produce:

• A first complete AI inventory

• A list of unknowns

• Clear system owners

• Vendor follow-up questions

• Product AI risk flags

• A short list of systems needing deeper review

After The Workshop

Within one week, clean up the inventory, confirm missing details, and assign follow-ups. The workshop is not the whole governance program. It is the map that shows you where to look next.