ai act checklist
AI Act checklist for companies
This checklist covers the gaps that show up most often in early AI projects.
In practice: In a first conversation we work out together which items are critical for you and what you can act on right away.
- Inventory the use cases and name the systems involved.
- Set a provisional risk class for each use case.
- Document data sources, consents and the GDPR legal basis.
- Clarify responsibilities (tech, legal, product, management).
- Define a monitoring and incident process for AI outputs.
- Prepare documentation that stands up in internal and external audits.
Trust signal: No standard pitch: you get a prioritized backlog of measures, not slides.
Get in touchWhat an enquiry sets off
- The reply comes from the engineer who would do the work, not from a sales layer that passes it on.
- A few sentences are enough for a first read. System access, architecture diagrams and an NDA come later, and only if this goes anywhere.
- No is a possible outcome. If the problem is solvable in-house, or there is no slot free right now, the reply says so.
Next step: look at the risk classes in detail or ask about a project start.