The European AI Act
Understand the European regulation on artificial intelligence and its implications for your organization.
Try for freeThe AI Act is the European regulation that governs the development and use of artificial intelligence systems within the European Union. Unlike a mere recommendation, it is a binding text, with sanctions that can reach several percent of global revenue for serious breaches.
Its scope extends well beyond AI system vendors: any organization that deploys or uses an AI system, including through third-party tools, is subject to its transparency and risk management requirements.
A classification by risk level
The AI Act classifies AI systems into four risk categories, with obligations proportionate to each level. An unacceptable-risk system is purely and simply banned; a high-risk system requires strengthened documentation and human oversight.
- Unacceptable risk: banned systems (social scoring, behavioral manipulation)
- High risk: automated recruitment, credit scoring, critical infrastructure
- Limited risk: transparency obligations toward users
- Minimal risk: no specific obligation
A staggered application timeline
The AI Act's obligations do not all take effect at once: the strictest bans apply first, followed progressively by heavier requirements for high-risk systems. Mapping your usage now avoids discovering a high-risk usage right when the obligation becomes effective.
How Kaliteq draws on it
Kaliteq's AI governance diagnostic helps map AI usage across the organization and assess its exposure against the AI Act's core principles, without substituting for formal legal advice on the precise classification of each usage.
Related Kaliteq diagnostics
Related glossary terms
Go further
Frequently asked questions
Does the AI Act apply to companies outside the EU?
Yes, as soon as they deploy an AI system whose output is used within the European Union, regardless of where the company is based.
Is using an off-the-shelf conversational assistant enough to be in scope?
Yes, using third-party generative AI tools in company processes falls within the scope of transparency requirements, even without in-house development.
What are the penalties for non-compliance?
They vary by severity, reaching several percent of global revenue for the most serious infringements, notably the use of banned systems.
Does the Kaliteq diagnostic guarantee AI Act compliance?
No, it helps map and prioritize at-risk usage, but formal compliance requires dedicated legal support for the most sensitive cases.
Ready to assess your organization?
Create your free account and start your first diagnostic in a few minutes.
Try for free