Transparency obligations apply in
— days
The AI Act will soon apply to your business
A chatbot on your site? ChatGPT in the office? Software that filters job applications? Then you have obligations — some of which apply already. Find out in 10 minutes where you stand.
No account needed · 10–15 min

60%+
of EU SMEs have not yet started
AI is hidden everywhere in your business — even where you least expect it
The AI Act is not just for tech companies. The law looks at what you use: the matching in your recruitment software, the chatbot in your web shop, the smart camera at the entrance, the ChatGPT tab employees keep open on the sly.
- ✓
Since 2 February 2025 AI training for staff is already a legal requirement — almost no one knows this.
- ✓
From 2 August 2026 the transparency obligations apply; the high-risk requirements are expected to be postponed to 2 December 2027 (not yet final).
- ✓
Up to €35 million or 7% of turnover fines can reach for the most serious infringements.
From "no idea" to a concrete action plan
No legal jargon, no consultancy project. Just recognisable questions about your day-to-day practice.
Answer recognisable questions
“Does your website have a chatbot?” and “Does software filter your job applications?” — we translate your answers into the law.
Your traffic-light verdict right away
Free: your complete AI inventory with a colour for each system — from green to red (action needed right away).
Full report + action plan
For €149 incl. VAT: obligations per system with legal articles and a priority list with deadlines. Delivered instantly as a PDF.
Free scan, complete report for €149
€149 incl. 21% VAT · one-off payment
An adviser quickly charges €2,000+ for the same inventory. The report is not legal advice, but it is exactly the framework you need — also when clients or supervisory authorities ask for it.
- FreeAI inventory: all the AI in your business mapped out
- FreeTraffic-light score per system and organisation
- €149Obligations per system, with exact legal articles
- €149Action plan by deadline: what to do now and what is coming towards 2026–2027
- €149Your legal role per system, clearly explained
- €149As a PDF — the first version of your AI records
Example verdict
This is what you see for free after the scan.
Emotion recognition in customer conversations
Prohibited practice
CV screening software
High risk
Chatbot on website
Transparency obligation
Internal ChatGPT use
Organisational obligation
In the full report
Per system: your concrete obligations, deadlines and exact legal articles.

Harry Stuurman
AI specialist & entrepreneur · HS Services
About me
I am Harry Stuurman. I have worked in IT for 25 years and have run my own company, HS Services, since 2008. In my day-to-day work as AI & EdTech Lead in international education I also do exactly what the AI Act now asks of every company: I wrote a Responsible AI policy, I assess together with a privacy officer which AI tools can be deployed responsibly, and I have trained over 500 professionals in safe and responsible use of AI.
It is precisely through that work that I notice how almost every organisation struggles with this. Not out of reluctance, but because no one has an overview. Companies often do not even know that they use AI — it is hidden in the recruitment software, the chatbot, the tools employees quietly work with. Let alone that they know there has been a legal training obligation since February 2025 for everyone who works with AI.
That combination — being able to write policy and to make it work on the shop floor — is exactly what SMEs need right now. That is why this scan exists: the questions a good adviser would ask, in plain language, at a price every SME can afford. So that you do not first find out where you stand when a client or supervisory authority asks.
The EU AI Act, in plain language
The AI Act (Regulation (EU) 2024/1689) is the first broad European regulation that specifically sets rules for both developing and using artificial intelligence. AI was already covered by other rules before, such as privacy legislation. The persistent misconception is that the law only applies to tech companies that build AI. That is not true: organisations that merely use AI fall under it too.
Does this also apply to my business?
The Act distinguishes various roles, including providers (who develop AI or place it on the market under their own name) and deployers (who use an AI system under their own responsibility in a professional capacity). If you use AI to write texts, assess candidates or answer customer questions, you are usually a deployer — the precise definition is in Article 3. Whether the law applies to you does not depend on your sector or size; sole traders and SMEs fall under it too. Note: the Act is risk-based. How heavy your obligations are depends on what the AI system does — from prohibited and high-risk to limited or minimal risk. Many everyday applications carry limited or no additional obligations, and for SMEs and start-ups the law moreover provides relief.
You are probably already using AI
AI is often embedded more deeply in everyday practice than you suspect: the chatbot on your website, the text and image generators employees use, software that pre-sorts job applications, or smart cameras in a shop or warehouse. Note: the same type of tool can, depending on its use, fall into very different categories — from a light transparency obligation to the strictest rules, and in rare cases even a ban. Which regime applies to your application depends heavily on the concrete use. The first step is therefore to get a clear picture of where AI sits in your business.
The timeline — and what recently changed
The Act is being introduced in phases. Since 2 February 2025 the duty of effort regarding AI literacy (Article 4) and the ban on certain practices (Article 5) have applied. On 2 August 2025 came, among other things, the rules for general-purpose AI models, the governance framework and the fines and enforcement regime. On 2 August 2026 the Act becomes largely applicable, including the transparency obligations (Article 50). The requirements for stand-alone high-risk AI (Annex III) were also set to take effect on 2 August 2026, but under a provisional EU agreement (the "Digital Omnibus", May 2026) they are expected to be postponed to 2 December 2027. That agreement has not yet been definitively adopted; the exact dates may therefore still change.
The fines rise in categories (Article 99). For the prohibited practices the highest ceiling applies: up to 35 million euros or 7% of total worldwide annual turnover, whichever is higher. For other infringements the maximums are lower (up to 15 million or 3%). For SMEs and start-ups, by contrast, the lower of those two amounts applies.
Why a scan instead of an expensive consultancy project
A consultancy project to map out where your business stands with AI can quickly cost hundreds to thousands of euros. The AI Act Wise provides a comparable initial inventory in about ten minutes, for a fixed amount you know in advance. You answer questions in plain language; the outcome is based on the text of the Act and gives a first indication of which elements may be relevant to your situation. It is a tool — not a full, tailor-made legal assessment — that gives you a clear first picture before you incur any costs.
This information is general in nature, reflects the state of affairs at the time of writing and does not constitute legal advice. Regulations and dates of application may change. The AI Act Wise is a self-assessment tool based on Regulation (EU) 2024/1689; for your specific situation, consult a specialist where necessary.
Frequently asked questions
›Does the AI Act also apply to SMEs?
Yes. Whether it applies does not depend on the size of your company — SMEs and organisations that merely use AI (deployers) fall under it too, not just companies that develop AI. The Act does, however, contain specific provisions in favour of SMEs and start-ups, such as a more favourable calculation of fines and priority access to regulatory sandboxes. Which obligations apply to you depends on how and for what you use AI — and that is exactly what the scan reveals.
›We only use ChatGPT or Copilot — do I need to do anything?
Probably yes. The Act does not look at the name of a tool, but at how it is used and the risk involved. Since 2 February 2025, everyone who deploys AI must ensure a sufficient level of AI literacy among staff (Article 4). How you use such tools also determines which further rules apply; the scan maps this out for your situation.
›What do I need to do right now?
Since 2 February 2025 two things apply: the duty of effort to ensure a sufficient level of AI literacy among staff who work with AI (Article 4), and a ban on certain prohibited AI practices (Article 5). A good first step is getting a clear picture of which AI your organisation uses and for what — after that you can see what specifically applies to you.
›What changes around 2 August 2026?
On that date the Act becomes largely applicable, including the transparency obligations (Article 50). The requirements for stand-alone high-risk AI (Annex III) were also set to take effect on 2 August 2026, but under a provisional EU agreement (the 'Digital Omnibus', May 2026) they are expected to be postponed to 2 December 2027. That agreement is not yet final; the exact dates may still change.
›What does non-compliance cost?
The Act has a tiered system of fines (Article 99). The highest tier — up to 35 million euros or 7% of total worldwide annual turnover — applies to the most serious infringements, the prohibited AI practices. For other infringements the maximums are lower (up to 15 million or 3%). For SMEs and start-ups, moreover, the lower of those two amounts applies.
›Do you also carry out the actions for me?
No. The scan and the report give you a clear roadmap: what is at stake, what you need to arrange and by when. Carrying it out — organising training, adding notices, adjusting processes — is up to you, or your own suppliers and advisers. That is what keeps the report affordable and independent.
›Is this legal advice?
No. The AI Act Wise is a self-assessment tool based on Regulation (EU) 2024/1689 and provides general information. For high-risk systems or in case of doubt, we recommend consulting a specialised lawyer; the report gives you exactly the right starting point for doing so.
›What happens to my data?
Your answers are used only to create your report. We do not sell data and we do not send spam. Payment is handled via Mollie (iDEAL, credit card). See the privacy statement for details.
›What does the report cost?
The report costs €149 including 21% VAT, a one-off payment. A comparable inventory by an adviser quickly runs to €2,000+, and doing nothing can cost you far more in the event of enforcement or a critical client. The free verdict already gives you an honest picture — the report is there for anyone who wants to document it and act on it. We charge 21% Dutch VAT on every order and do not apply VAT reverse-charge (intra-Community 0%).
Know where you stand within 10 minutes
Transparency obligations apply in — days. The free scan gives you your AI inventory and traffic-light verdict right away.
Start the free scan →