EU AI Act Readiness

What it does

Assessing EU AI Act compliance for AI systems before EU deployment or during annual compliance refresh.

Procedure

When this skill is activated, Chalie follows these steps:

1. Use memory to recall any prior AI Act assessments or compliance decisions for the system in scope.
2. Confirm the AI system’s description using memory to surface any stored context: what it does, who uses it, where it is deployed, and the organization’s role (provider, deployer, importer, or distributor).
3. Screen for Article 5 prohibited practices — use document to record findings as you go; if any of the 8 prohibited categories apply, stop and note that the system cannot be placed on the EU market.
4. Classify the system’s risk tier under Article 6 and Annex III using search to confirm current Annex III category boundaries: minimal risk, limited risk, high-risk, or GPAI — note that the Annex III carve-out does not apply if the system profiles natural persons.
5. For high-risk systems, use document to determine and record whether conformity assessment requires Module A (internal control with harmonized standards) or Module H (full QMS with notified body, required for biometrics).
6. Check Article 50 transparency obligations and use document to record the assessment: AI interaction disclosure, synthetic content marking, emotion recognition disclosure, and deepfake labeling — these obligations are in force from 2 August 2025.
7. Use search to verify whether any applicable harmonized standards have been published that enable presumption of conformity.
8. Use document to save a compliance readiness report listing each question, the verdict, and the outstanding actions required before EU deployment.

Version

v1 (curated)