Deeploy launched the EU AI Act Hub, a free, continuously updated knowledge base that explains the EU AI Act in plain language. It offers a risk self-assessment tool, a full compliance timeline from 2024 to 2030, deep-dive articles across nine topics, and a glossary and FAQ, helping compliance teams, AI leads and executives understand which obligations apply to their AI systems and when.
Understanding the EU AI Act shouldn't require a law degree
The EU AI Act is the world’s first comprehensive law on artificial intelligence, and it is not a simple read. It runs to more than 150 pages and 113 articles, sorts AI into four risk tiers, assigns different duties depending on whether you build, deploy, import or distribute a system, and phases in obligations across several years. On top of that, the deadlines themselves keep moving: the Digital Omnibus, finalised in mid-2026, shifted several high-risk deadlines to 2027 and 2028 while leaving the transparency and general-purpose AI obligations on 2 August 2026.
For the teams who actually have to comply, that combination of density and change is the real problem. Most explainers either oversimplify the Act or bury the reader in legal text, and many of the summaries circulating online still describe last year’s version of the timeline.
To help, we built the EU AI Act Hub: a continuously updated, plain-language knowledge base for anyone who needs a clear, accurate overview of what the Act requires and when.
Contents of the EU AI Act Hub
The Hub is organised around the questions we hear most often from compliance leads, AI teams and the executives who answer to them:
- A risk self-assessment tool: Answer a short set of questions and find out which of the Act’s four risk tiers your AI system falls into, the starting point for every other obligation.
- A full compliance timeline: Every relevant deadline from 2024 through 2030, kept current as the rules change, so you are never planning against an outdated date.
- Deep-dive articles across nine topics: Practical explanations of the areas teams struggle with most, from high-risk requirements to your responsibilities when building on someone else’s foundation model.
- A glossary and FAQ: Clear definitions and direct answers for the moments when you need one fact quickly rather than a long read.
Why we keep it updated
AI regulation is not static. Deadlines shift, guidance is published, and enforcement practice develops. A reference that was accurate six months ago can quietly mislead a team today. We maintain the Hub as the landscape moves so that the information you act on reflects the Act as it stands now, not as it stood when the summary was first written.
Reading about compliance is the first step. Operating it, proving in production that your AI systems do what your documentation says, is the harder and more important work, and it is what the Deeploy platform is built for.
Explore the EU AI Act
Start with the risk self-assessment, follow the timeline, or browse the deep-dive articles:
About the author: Maarten Stolk is the co-founder and CEO of Deeploy, the leading European AI governance platform, and the co-founder of AI agency Enjins and national AI hub The Stack in Amsterdam.
Want more? Download our whitepaper.
We dive into this and much more in the complete whitepaper. Co-developed with 12+ industry partners, including Deloitte & BearingPoint.
The question isn’t whether you’ll need robust AI governance. It’s whether you’ll build it in time.
