Definition
The world's first regulatory framework for artificial intelligence, adopted by the European Union in 2024, with progressive application by risk level.
The AI Act classifies AI systems by risk level (unacceptable, high risk, limited risk, minimal risk) and imposes proportionate obligations: a ban on unacceptable practices since February 2025, transparency for general-purpose models since August 2025, and reinforced obligations for high-risk systems.
The timeline has shifted: the “Digital Omnibus” package adopted in June 2026 pushed most high-risk obligations back to December 2027. Companies deploying AI internally remain mainly concerned by obligations of transparency, documentation and human oversight.
An SME uses an AI tool to screen CVs. This use is classed as high-risk: it requires human oversight, documentation and informing candidates. An assistant that summarises internal documents, on the other hand, falls under limited risk.
As soon as you deploy or buy an AI tool: you need to know which risk category it falls into. Uses in recruitment, credit, healthcare or education are the most tightly regulated.
Does the AI Act apply to SMEs?
Yes. It applies to any organisation that provides or uses an AI system in the European Union, whatever its size. The obligations depend on the risk level of the use, not on the size of the company.
All AI models, deployed on your infrastructure, with your documents connected.