Prohibited AI practices are now regulated by the European AI Act, in force since August 2024.
Regulation (EU) 2024/1689 (AI Act), in force since August 2024, imposes a strict framework on artificial intelligence systems in Europe. Among the key measures, Article 5 prohibits AI practices deemed unacceptable due to the major risks they pose to fundamental rights and safety.
Prohibited AI practices under Article 5 of the AI Act
💡 What are the prohibited practices?
➡ Manipulation and exploitation of vulnerabilities
Systems using subliminal, manipulative or deceptive techniques to alter individuals’ behaviour and decision-making, in particular by exploiting weaknesses related to age, disability or socio-economic precarity.
➡ Social scoring
The evaluation of individuals based on their social behaviour or personal characteristics leading to unjustified or disproportionate treatment.
➡ Criminal risk prediction
Systems assessing the likelihood that a person will commit an offence, based solely on profiling or personal characteristics.
➡ Mass scraping of facial images
The creation of facial recognition databases through automated collection of images from the Internet or video surveillance.
➡ Emotion recognition in the workplace and at school
The analysis of emotions of employees and students, except for medical or safety reasons.
➡ Biometric categorisation based on sensitive criteria
The use of biometric data to infer race, political opinions, religious beliefs or sexual orientation.
➡ Real-time remote biometric identification
The use of remote biometric identification systems in public spaces for law enforcement purposes, except under strictly regulated exceptions.
🔎 What are the risks?
Companies and organisations that violate these prohibitions face penalties of up to EUR 35 million or 7% of global annual turnover. Market surveillance authorities will be responsible for enforcing these rules.
💬 What impact on AI stakeholders?
This framework requires increased vigilance from AI system developers and users. It strengthens the responsibility of providers and deployers throughout the AI lifecycle, while ensuring innovation that respects fundamental rights.
🌍 Responsible AI is now a legal obligation in Europe!
Companies, are you ready to ensure the compliance of your AI systems? 🤖⚖️
Want to learn more:
https://lawgitech.eu/category/intelligence-artificielle/
Here are the guidelines:
Définition du système d’intelligence artificielle (IA) selon l’AI Act