You are probably already using AI in your business. For text, emails, invoices, customer questions or analysis. From 2 August 2026, key obligations under the EU AI Act become more concrete, especially around high-risk AI and transparency. The law applies not only to those who build AI, but also to those who use AI professionally.
Many self-employed and entrepreneurs still think of AI as standalone tools like ChatGPT, Claude, Gemini, Copilot or Canva AI. In practice, AI is now often hidden in software you use every day: Outlook, your CRM, accounting software, video meetings, website tools, marketing platforms, project management.
Integrations count too. The moment you connect Gmail, Google Drive, Notion, Slack or your company documents to ChatGPT, Claude or an AI agent, you are in practice using an AI system. That is exactly why an inventory matters so much: you map not only the tools you deliberately chose, but also the AI that is already woven into your daily operations.
The better question: what are we using AI for, with which data, and with what impact on people?
The EU AI Act classifies AI systems by risk. You look not only at the tool, but above all at how it is used.
Most everyday AI use: text, ideas, emails, summarising meetings, recognising invoices. The impact stays low as long as AI supports and does not decide.
Situations where people need to know they are dealing with AI: a chatbot on your website, AI-generated images, automated customer communication.
Applications that affect decisions about people: AI that screens applicants, evaluates employees, scores customers or helps assess creditworthiness. Here, "we use a well-known tool" is not enough.
Manipulative AI, social scoring or applications that exploit vulnerable groups are not allowed. The first prohibited practices have applied since February 2025.
Do not start with a big compliance project. For each tool, map out:
The Belgian FPS Economy offers a free tool for this: aiactvoormij.be. It works mainly from the legal framework. That is why Dorien made a practical version: a workable document for self-employed and entrepreneurs, and their teams.
In line with the requirements, translated into a workable document you can use to map your AI use in a team meeting. No obligation.
For each tool, fill in the purpose, data, risk, transparency and oversight. Ready for your team.
Download ExcelThe background and approach in PDF. Ideal to go through with your team.
Download PDFBased on Regulation (EU) 2024/1689. This is a practical tool, not legal advice.
The EU AI Act does not ask every entrepreneur to become a lawyer. It asks that you know which AI you use, why, and where the impact sits.
No panic project. Just professional AI use. Need help getting this in order for your business?
Book a callThe regulation entered into force in August 2024. The first prohibited practices have applied since February 2025. Key obligations around high-risk AI and transparency become concrete from 2 August 2026.
Most likely yes. The law applies not only to those who build AI, but also to those who use AI professionally. AI is often hidden in software you use daily: Outlook, your CRM, accounting software, marketing platforms.
Do not start with a large compliance project, but with an AI inventory. Map per tool: purpose, user, data, risk, transparency, oversight and owner.
No. This is a practical tool to map your AI use, in line with Regulation (EU) 2024/1689. For your specific situation you can always book a conversation.