The European AI Act: what changes for businesses and IT professionals?
The European Union has introduced the AI Act (EU Regulation 2024/1689), the world's first binding legal framework designed to regulate the development and use of Artificial Intelligence. For SMEs and system integrators, compliance is no longer an option but a necessity to avoid penalties and ensure the conformity of their systems.
The regulation takes a risk-based approach, dividing technologies into
4 Risk Bands:
Unacceptable risk (Prohibited): Systems such as social scoring or behavioral manipulation.
High risk: Software used in critical areas such as personnel selection, healthcare or biometrics, which require strict certifications.
Limited risk: Chatbots and generative AI, for which a strict transparency obligation toward the user applies (including watermarks on synthetic content).
Minimal or no risk: Spam filters and video games, exempt from heavy requirements.
The immediate practical steps for businesses:
Tool mapping: Catalog all the AI software in use, checking its risk category.
Transparency and Watermarking: Clearly declare the use of chatbots and of content generated or manipulated by AI.
Training (AI Literacy): Ensure adequate digital and AI literacy for company staff.
Data Governance: Ensure that data processing complies with company security and the requirements of the GDPR.
👉 Want to dig into all the deadlines and technical details?
Download and read the full PDF attachment to have the summary guide and the official sources at your fingertips.
