European Council Grants Final Approval to Groundbreaking AI Act

European Council Grants Final Approval to Groundbreaking AI Act

In a landmark decision on Tuesday, ministers from the European Union (EU) member states have given their final approval to the bloc’s comprehensive Artificial Intelligence (AI) Act.

This pioneering legislation adopts a risk-based approach, aiming to regulate AI technologies while fostering innovation and protecting fundamental rights within the EU.

The AI Act is designed to address the diverse applications of AI with varying levels of regulatory scrutiny based on associated risks. High-risk AI systems, such as those used in critical infrastructure, education, and law enforcement, will be subject to stringent regulations. These include rigorous testing, documentation requirements, and ongoing compliance checks to ensure safety and ethical standards are met.

Among the key provisions, the AI Act bans systems involved in cognitive behavioral manipulation and social scoring. This prohibition reflects the EU’s commitment to safeguarding individual autonomy and dignity. Additionally, the Act outlaws predictive policing techniques based on profiling and any AI systems that use biometric data to categorize individuals by sensitive attributes such as race, religion, or sexual orientation. These measures aim to prevent discrimination and uphold the EU’s values of equality and human rights.

For general-purpose AI systems, the Act mandates compliance with transparency requirements unless these systems present systemic risks, in which case stricter rules apply. These transparency measures are intended to ensure that AI applications are understandable and accountable, allowing users and stakeholders to have clear information about their functionality and limitations.

However, the Act exempts AI systems used exclusively for military, defense, and research purposes, recognizing the unique needs and security considerations of these domains.

To oversee the implementation and enforcement of the AI Act, several governing bodies will be established. An AI office will serve as the central regulatory authority, while a scientific panel of experts will provide technical guidance. Additionally, an AI board comprising representatives from member states will coordinate enforcement activities, and an advisory forum will facilitate stakeholder engagement, ensuring a diverse array of perspectives inform the regulatory process.

The AI Act will come into effect 20 days after its publication in the EU’s Official Journal, following its signing by the presidents of the European Parliament and the European Council. This historic legislation positions the EU at the forefront of global AI governance, setting a precedent for balancing technological advancement with ethical responsibility.

With the AI Act, the EU aims to create a framework that not only regulates the risks associated with AI but also promotes innovation and trust in AI technologies, ensuring they contribute positively to society.