How-to
Which EU AI Act rules apply to you? Take the 2-minute check
KL
Klaar
6 July 2026 · 5 min read
6 July 2026 · 5 min read
Try it here
1 / 1
Which of these are you?
Pick every role that fits. You can be more than one, for example a deployer and a model maker.
The EU AI Act has a reputation for being a moving target, and in 2026 it moved again. If you run a small European team, the honest question is not whether you are affected, because you almost certainly are, but which parts apply and by when. The check above answers exactly that in about two minutes. Here is what sits behind it.
What changed in 2026
In May 2026 the Council and Parliament agreed the Digital Omnibus, a package that simplifies and, in places, delays the AI Act. The headline change for most businesses is that the high-risk obligations for Annex III systems, the heavy duties around hiring tools, credit scoring and similar decisions, moved from 2 August 2026 to 2 December 2027. High-risk AI embedded in regulated products moved to 2 August 2028.
What did not move matters just as much:
- AI literacy (Article 4) and the ban on prohibited practices (Article 5) have applied since 2 February 2025. They are live today, for everyone, regardless of size.
- Transparency (Article 50) still applies from 2 August 2026. If people talk to your chatbot or you publish AI-generated content, you have to be clear about it.
- General-purpose AI model rules have applied since 2 August 2025.
- New prohibitions, covering non-consensual intimate imagery and child sexual abuse material, and the end of the AI-content watermarking grace period, land on 2 December 2026.
The delay is genuine relief on high-risk, but it is not a pause. The duties that are already live, literacy and prohibited practices, are the ones a small team is most likely to trip on.
What the check tells you
The check maps your answers to the specific articles in scope for you, whether you are a deployer who uses AI, a provider who builds it, or both. For each one it shows whether the duty is live now, coming soon, or further out, and it places you on the real timeline with a "we are here" marker so you can see how long you actually have.
What to do with your result
Most small teams land in a similar place: AI literacy and a light usage policy now, transparency by August 2026 if you are public-facing, and high-risk work only if you screen candidates or make automated decisions about people. The value is in knowing which of those is yours, in what order to tackle them, and keeping dated evidence of the effort.
At the end of the check you can leave your email for a step-by-step outline written around your exact answers. And if you would rather just get it done, Klaar turns the whole thing into an afternoon: a risk-classified tool inventory, role-based training, a usage policy and a dated compliance record you can hand to anyone who asks.
Turn this into a finished record
Klaar walks you through inventory, training, policy and the dated evidence.
Keep reading
