European Parliament ready to negotiate first-ever rules for safe and transparent AI

← All articles
Artificial intelligence · EN

European Parliament ready to negotiate first-ever rules for safe and transparent AI

The European Parliament adopted its AI Act negotiating position with 499 votes in favor, 28 against, and 93 abstentions. The rules aim to ensure EU-aligned AI development, prioritizing human oversight, safety, privacy, transparency, non-discrimination, and social-environmental wellbeing.

Before the law is passed, details will be negotiated in a trialogue between the Parliament, the EU Commission and the Council, so there is still the possibility to make improvements.

Here a summary:

🚫 Prohibited AI practices:
 - Ban on real-time and post remote biometric identification in public spaces
 - Ban on biometric categorization using sensitive characteristics
 - Ban on predictive policing systems
 - Ban on emotion recognition systems in various sectors
 - Ban on untargeted scraping of facial images for facial recognition databases

📱 High-risk AI additions:
 - Inclusion of AI systems influencing elections and social media recommender systems

⚖️ Obligations for general purpose AI:
 - Risk assessment and registration in EU database for foundation models
 - Transparency requirements and safeguards against generating illegal content
 - Public availability of detailed training data summaries

🗣️ Support for innovation and citizen rights:
 - Exemptions for research activities and AI components under open-source licenses
 - Establishment of regulatory sandboxes for testing AI
 - Citizens’ right to file complaints and receive explanations
 - Reformed role of EU AI Office for monitoring implementation

#ai #surveillance #chatgpt

https://lnkd.in/eWBut\_AQ

Discuss on LinkedIn