Every AI feature in your own product falls under one of the EU AI Act’s four risk tiers, whether prohibited, high-risk, subject to a transparency duty, or minimal, and not knowing your own classification makes it easy to miss a required disclosure.
Every AI-touching feature gets classified, the reasoning gets documented rather than just the label, and the Article 50 disclosure text is ready for the consent screen or the Terms of Service: one always-visible sentence plus a collapsible detail block. Added to that is wording on human oversight, which satisfies the matching DSGVO requirement at the same time.
Two checklists come with it: a plain-language one for everyone involved, and a shorter one for management sign-off.
What ports as-is
- The classification into the four tiers: prohibited, high-risk, subject to transparency duty, or minimal
- Two check questions decide the tier: does an AI make an autonomous decision with legal or significant effect on a person, and is human oversight built into the product itself rather than just claimed on the surface
- The two-tier Article 50 disclosure text: one visible sentence plus a collapsible detail block
- Ready-made Terms of Service wording for the use of AI
- An overview of which EU AI Act deadlines apply to this product and which do not
- A plain-language checklist for everyone involved
- A shorter checklist for management sign-off
What we build for you
- The business's own list of AI-touching features and the model or provider behind each one
- The disclosure wording in the business's own language and register
- The surface the text appears on, whether the consent screen or the Terms of Service
- Legal review before any statement is used externally as proof of compliance
