AI Act: first prohibitions apply
2025-02-02 · Publisher : Union européenne · Domain : Artificial Intelligence
Six months after its entry into force, the EU AI Act starts to apply: practices deemed unacceptable are banned and organizations must ensure AI literacy among their staff.
What changes
- Since February 2, 2025, the practices listed in Article 5 are banned throughout the EU.
- Targets include social scoring, harmful manipulation and emotion recognition at work and in schools.
- Article 4 requires providers and deployers to ensure a sufficient level of AI literacy among their staff.
- Fines can reach €35 million or 7% of worldwide turnover; the penalty regime applies from August 2025.
- On February 4, the Commission publishes non-binding guidelines on prohibited practices.
Having entered into force on August 1, 2024, Regulation (EU) 2024/1689 on artificial intelligence applies in stages. February 2, 2025 marks the first one: Chapters I and II become applicable, with the general provisions, the AI literacy obligation and the list of prohibited practices. ### What happened - **Prohibited practices (Article 5)**: subliminal or manipulative techniques causing significant harm, exploitation of vulnerabilities related to age, disability or social situation, social scoring, assessment of offending risk based solely on profiling, building facial recognition databases through untargeted scraping of images, emotion recognition in the workplace and in education (except for medical or safety reasons), biometric categorization to infer sensitive data, and real-time remote biometric identification in public spaces for law enforcement, save for strictly limited exceptions. - **AI literacy (Article 4)**: providers and deployers must take measures so that their staff have a sufficient level of knowledge of the systems they use. - **Guidelines**: on February 4, 2025, the Commission publishes its guidelines on prohibited practices. They are non-binding; authoritative interpretation rests with the Court of Justice of the EU. ### Key figures | Item | Value | |---|---| | Entry into force of the regulation | August 1, 2024 | | Prohibitions apply | February 2, 2025 | | Maximum fine (prohibited practices) | €35 million or 7% of worldwide turnover | | Penalty regime applies | August 2, 2025 | ### Why it matters For the first time, uses of AI are banned by law across a market of 450 million people. For an IT services company the effect is twofold: checking that no client project falls under Article 5, and training teams, since the AI literacy obligation applies to every organization deploying these systems. ### Key takeaways The AI Act is no longer a future text: its prohibitions apply since February 2, 2025, ahead of the obligations on general-purpose models (August 2025) and on high-risk systems.
Links
Tags : IA, Régulation, AI Act, Europe