An EU AI Act compliance audit
that classifies your real systems.
The EU AI Act assigns obligations by risk tier, and the first job is knowing where each of your systems falls. We audit your AI portfolio against the Act, classify each use case by risk, name the obligations it triggers, and produce the documented risk register your legal counsel can sign off.
Where each system falls on the risk scale
We inventory the AI systems you build and use, then classify each against the Act's tiers: prohibited, high-risk, limited-risk with transparency duties, and minimal-risk, including the general-purpose and generative-AI obligations where they apply.
Classification is the load-bearing step. Getting a system's tier wrong is how you either ship something you should not or smother a useful system in obligations it never triggered.
What each tier actually requires
For each classified system we name the concrete obligations: documentation, transparency notices, human oversight, data governance, and the records the Act expects you to keep. We separate genuine requirements from over-caution that quietly kills useful work.
The result is practical: not a treatise on the regulation, but a list of what you must do, per system, to be inside the lines.
A risk register your counsel can sign
You leave with a documented risk register: each system, its classification, the obligations it triggers, and the gaps to close, written so your own legal counsel can review and sign off on the legal position.
This is the EU AI Act slice of the full Audit & Consulting engagement. For the team understanding behind it, the EU AI Act Training brings your people to a shared baseline.
Common questions.
Direct answers to the questions we get asked the most. If yours isn't covered, write to the team.
Know where your AI falls under the EU AI Act.
Tell us the AI systems you run. We classify each by risk, map the obligations, and hand you a register your counsel can sign.