The AI Act: what really applies to you, and how to comply.
If you use chatbots, automation or AI-generated content, we map the uses, classify them by risk and implement transparency and documentation. Without the panic.
Sound familiar?
And what staying unsure costs you.
You use a chatbot on your site and don't know if you must disclose it.
People writing to it think they're talking to a person. When they find out they weren't, you're the one who loses their trust, not whoever sold you the chatbot.
You generate text or images with AI and don't know if they must be labelled.
Without a rule, everyone in the company does it their own way. The day you need to show what you labelled, there's nothing to show.
You hear about fines and deadlines, but don't know where to start.
You put it off until a real deadline arrives. Then you decide in a rush, and often buy more than you need.
What we do
Mapping and classification
Where you use AI in products and processes, and which risk category each use falls into.
Transparency in the product
Notices for chatbots and AI interactions, and labels on generated content, built into the site or app.
Also
Documentation
The record the regulation asks for.
GDPR alignment
Automated decisions and data.
Compliance plan
Priorities and steps, in order.
We start with ourselves
On this site, a model reads comments and flags them. A person decides, and we say so.
Our own site
A page says where the model runs, what we send it and who decides. It's the only place on the site where AI is used.
Why not online guides, or a lawyer?
They're fair choices, and sometimes they're enough. Here's what changes when the people looking for AI in your tools are the ones who build software.
| What matters | DIY, with online guides | Consultant or law firm | M's Works |
|---|---|---|---|
| What it looks at | The regulation in general, not your case. | Contracts, and what you tell them about your tools. | Where a model actually runs, in the site, the app or your processes. |
| Notices and labels | You write them and put them in yourself. | Tells you what to write. Putting them into the product is up to whoever builds it. | We build them into the site or app. |
| Legal questions | You interpret them on your own. | That's their job: they give opinions and sign off. | We don't give legal advice. Where you need it, we say so. |
| When the timetable changes | You notice if you follow the news. | They keep you posted, if it's part of the engagement. | We keep you up to date on deadlines and the regulation. |
What it looks at
M's Works: Where a model actually runs, in the site, the app or your processes.
DIY, with online guides: The regulation in general, not your case.
Consultant or law firm: Contracts, and what you tell them about your tools.
Notices and labels
M's Works: We build them into the site or app.
DIY, with online guides: You write them and put them in yourself.
Consultant or law firm: Tells you what to write. Putting them into the product is up to whoever builds it.
Legal questions
M's Works: We don't give legal advice. Where you need it, we say so.
DIY, with online guides: You interpret them on your own.
Consultant or law firm: That's their job: they give opinions and sign off.
When the timetable changes
M's Works: We keep you up to date on deadlines and the regulation.
DIY, with online guides: You notice if you follow the news.
Consultant or law firm: They keep you posted, if it's part of the engagement.
The first step is a free check.
Tell us where you use AI, on your site or in day-to-day work. In the preliminary check we look at what really applies to you and what doesn't.
- A reply within 24 hours.
- A first call and a preliminary check, both free.
- A list of your AI uses, and what doesn't apply to you, written down too.
- If needed, the full audit: risk category and duties, use by use.
- If you then give us the work, the audit isn't billed separately.
The first call and the preliminary check are always free. The full audit has a cost, which you only pay if you don't go ahead or give the work to someone else.
- 01AssessmentWe find and map all your uses of AI.
- 02ClassificationThe risk level and the duties that follow.
- 03MeasuresTransparency, documentation and GDPR alignment.
- 04OngoingWe keep you up to date on deadlines and the regulation.
you use a chatbot, generate content with AI or are bringing it into your processes, and want to know what you actually have to do, without the panic.
you want a signed legal opinion. That comes from a lawyer, and we'll tell you when you need one.
Before starting, people ask us
“Does it really apply to me?”
We'll tell you, separating real duties from the noise, use by use.
On our site, a model reads comments. We classified it: this use isn't covered by the AI Act's transparency duties, and the page says so. We declare it anyway, by choice.
How we use AI →“The deadlines keep changing.”
That's why we don't chase dates: you know your category and have the measures ready.
“Is it just paperwork?”
No: we build notices and labels into the product, not just into a document.
On our site, the model can't throw a comment away. Whatever it doesn't pass waits for a person, and nothing is deleted automatically. The rule lives in the code.
What people ask us
No. We implement the technical measures and, where legal advice is needed, we say so.
The timeline has already been revised more than once and will keep moving. What helps is knowing which category you fall into and having transparency and documentation ready.
Documentation and transparency duties: saying when content or an interaction is AI-generated.
Does the AI Act apply to you?
We'll check your AI uses and tell you which obligations you have and how to meet them.