AI Act compliance assessment
Welcome
This 22-question questionnaire gives a first picture of your practices, based on your own statements.
Estimated duration: 10 minutes
1. What is your company's average annual turnover?
This information is used only to estimate potential penalties in case of non-compliance.
2. What is your industry?
You may select several industries if applicable.
3. How do you use AI?
You may select several types of use if applicable.
4. Do you have technical documentation for your AI solution?
5. Does your AI comply with transparency and bias-control rules?
6. Do you have ways to check that your AI treats everyone fairly?
7. Do you have controls ensuring customer data is processed transparently?
8. Do you track the carbon footprint or energy use of your AI solutions?
9. Do you have measures to limit the environmental impact of your AI?
10. If so, briefly describe your actions to limit the carbon footprint
This question is optional if you answered 'No' to the previous one.
11. Can your models or data easily be moved to another provider?
12. Do you have a policy to prevent vendor lock-in?
13. Explain simply how you handle portability.
This question helps us understand your approach to data portability.
14. Does your AI system process personal data?
15. If so, have you identified a legal basis for this processing?
16. Have you carried out or planned a data protection impact assessment (DPIA)?
17. Are your teams trained in, or made aware of, the responsible use of AI?
18. Do you have an internal policy governing the development or use of AI (ethics charter)?
19. Do you have a documented risk management system, kept up to date throughout the AI system's lifecycle (Art. 9)?
A planned, iterative process: risk identification, mitigation measures and regular review.
20. Have the system's residual risks been assessed and judged acceptable, with the corresponding justification (Art. 9)?
Once mitigation measures are applied, each residual risk must be judged acceptable and communicated to the deployer.
21. Does the system automatically record events occurring during its operation (logs, Art. 12)?
Logging must make it possible to identify risk situations after the fact and to monitor how the system operates.
22. Are people exposed to the system informed that they are interacting with AI, or that content is AI-generated (Art. 50)?
An obligation applicable since 2 August 2026, regardless of risk level.