A free CSV template for building your first AI inventory — the documented list of every AI system your business uses. Download it, fill one row per tool, and you have the starting point every AI Act assessment needs.
Download the template
ai-inventory-template.csv → Opens in Excel, Google Sheets, or Numbers. Includes one filled-in example row per category so you can see the level of detail expected.
What each column is for
Tool name
The actual product name — specific enough that two people reading the row agree what it is.
Category
Chat assistant, hiring, finance, embedded feature, etc. — groups tools by the kind of risk they carry.
Owner
The team or person who can answer questions about it. Without this, follow-up dies.
Purpose
What it actually does, in one line — not marketing copy.
Data types processed
Determines a lot of what's at stake if it's misused or breached.
Who it affects
Customers, employees, or candidates — this is what decides whether disclosure obligations apply.
Provider
Who you'd need to contact, and whose terms govern the data.
Jurisdictions where used
Different rulesets apply depending on where the affected people are.
Annex III / high-risk?
Flags tools needing the closest attention under the EU AI Act.
Disclosed to affected people?
The literal gap the EU AI Act's Article 50 transparency duty is about.
Categories to check before you think you're done
The tools people forget are usually the ones bundled inside software they already use, not the ones they went out and bought.
- Chat assistants & writing tools: ChatGPT, Microsoft 365 Copilot, Google Gemini, Claude, Grammarly
- Customer-facing chatbots & voice bots: Website chat widget, Intercom Fin, Zendesk AI agents, AI phone / voice agent
- AI features inside software you already use: Salesforce Einstein, HubSpot AI, Notion AI, Zoom AI Companion, Canva Magic Studio
- Hiring & HR: CV / résumé screening, Video-interview analysis, Candidate ranking in your ATS, Performance analytics
- Finance, credit & risk: Credit scoring, Loan or BNPL approval, Fraud detection, Invoice / expense extraction
- Marketing & generated content: Image generators (Midjourney, DALL·E), AI copywriting, Product recommendations, Ad targeting
- Coding & internal tools: GitHub Copilot, Cursor, An LLM API call inside your own product, In-house ML models
Frequently asked questions
Is a spreadsheet enough, or do I need software?
A spreadsheet is a legitimate starting point and better than nothing — the template below gets you moving today. Its limits show up over time: nobody updates it when a vendor adds a new AI feature, and it won't score each tool against the actual regulation for you. Treat it as step one, not the finished system.
Who should fill this out?
Ask each team lead directly what they use, rather than relying on a central IT list — most AI adoption happens outside procurement. See our piece on shadow AI for why the tools nobody officially approved are usually the ones missing from a first pass.
How often should I update it?
At minimum, whenever you adopt a new tool or a vendor announces a new AI feature in something you already use. A quarterly re-ask of each team catches what slips through in between.
Skip the spreadsheet maintenance
DiscloseAI's free assessment builds the same inventory as a guided flow and scores each tool against the EU AI Act automatically — no manual upkeep. Free EU AI Act check →