Artificial intelligence (AI) act: Council gives final green light to the first worldwide rules on AI

🌈 Abstract

The article discusses the European Union's landmark legislation, the Artificial Intelligence (AI) Act, which aims to harmonize rules on AI across the EU. The law takes a risk-based approach, with stricter rules for high-risk AI systems that could cause harm to society. The AI Act is the first of its kind in the world and aims to foster the development and uptake of safe and trustworthy AI systems, while ensuring respect for fundamental rights and stimulating innovation in Europe.

🙋 Q&A

[01] Classification of AI systems as high-risk and prohibited AI practices

1. What are the key aspects of the classification of AI systems under the new law?

• The new law categorizes different types of AI systems according to risk:
  • AI systems presenting only limited risk are subject to light transparency obligations
  • High-risk AI systems are authorized but subject to a set of requirements and obligations to gain access to the EU market
  • AI systems such as cognitive behavioral manipulation and social scoring are banned from the EU as their risk is deemed unacceptable
  • The use of AI for predictive policing based on profiling and systems that use biometric data to categorize people by race, religion, or sexual orientation are also prohibited

[02] General purpose AI models

1. How does the AI Act address the use of general-purpose AI (GPAI) models?

• The AI Act distinguishes between GPAI models:
  • GPAI models not posing systemic risks are subject to some limited requirements, such as transparency
  • GPAI models with systemic risks have to comply with stricter rules

[03] A new governance architecture

1. What are the key governing bodies established under the AI Act to ensure proper enforcement?

• An AI Office within the Commission to enforce the common rules across the EU
• A scientific panel of independent experts to support the enforcement activities
• An AI Board with member states' representatives to advise and assist the Commission and member states on consistent and effective application of the AI Act
• An advisory forum for stakeholders to provide technical expertise to the AI Board and the Commission

2. What are the penalties for infringements to the AI Act?

• Fines are set as a percentage of the offending company's global annual turnover in the previous financial year or a predetermined amount, whichever is higher
• SMEs and start-ups are subject to proportional administrative fines

[04] Transparency and protection of fundamental rights

1. What are the key transparency and fundamental rights requirements under the AI Act?

• Before deploying high-risk AI systems, a fundamental rights impact assessment must be conducted
• High-risk AI systems and certain users of such systems (public entities) must be registered in the EU database for high-risk AI systems
• Users of emotion recognition systems must inform natural persons when they are being exposed to such a system

[05] Measures in support of innovation

1. How does the AI Act promote innovation?

• The AI Act provides for an innovation-friendly legal framework and aims to promote evidence-based regulatory learning
• The new law foresees the establishment of AI regulatory sandboxes, enabling a controlled environment for the development, testing, and validation of innovative AI systems, including in real-world conditions

[06] Next steps

1. What are the next steps for the implementation of the AI Act?

• After being signed by the European Parliament and Council, the AI Act will be published in the EU's Official Journal and enter into force 20 days later
• The new regulation will apply two years after its entry into force, with some exceptions for specific provisions