The EU Council has approved the AI Act, creating the first global regulatory framework for artificial intelligence, focusing on risk-based classification and stringent rules for high-risk systems. Photo: Shutterstock

The EU Council has approved the AI Act, creating the first global regulatory framework for artificial intelligence, focusing on risk-based classification and stringent rules for high-risk systems. Photo: Shutterstock

The AI Act, approved by the European Council on Tuesday, establishes comprehensive regulations for artificial intelligence, mandating transparency, enforcing strict regulations for high-risk AI systems and creating a new governance structure.

The council of the European Union has officially the artificial Intelligence (AI) Act on Tuesday 21 May 2024, a landmark legislation designed to harmonise rules on AI within the EU. This pioneering law, which follows a “risk-based” approach, aims to set a global standard for AI regulation. Its approval marks a significant milestone in AI governance, setting forth comprehensive regulations to ensure the development and adoption of safe and trustworthy AI systems.

Classification and regulation

The AI Act introduces a tiered classification system based on the risk level posed by different AI systems. AI systems deemed to present limited risk will face minimal transparency obligations. Conversely, high-risk AI systems will be allowed access to the EU market only if they meet stringent requirements and obligations. Systems engaging in practices such as cognitive behavioural manipulation and social scoring will be banned due to their unacceptable risk levels. Additionally, the law prohibits the use of AI for predictive policing based on profiling and for systems using biometric data to categorise individuals by race, religion or sexual orientation.

General purpose AI models

The legislation also addresses general-purpose AI (GPAI) models. Those not posing systemic risks will be subject to limited requirements, primarily concerning transparency. However, GPAI models identified as having systemic risks will need to adhere to stricter regulatory standards.

Governance and enforcement

To ensure effective enforcement of the AI Act, a new governance framework has been established. This includes the creation of an AI Office within the European Commission, tasked with enforcing the regulations across the EU. A scientific panel of independent experts will support enforcement activities, while an AI board comprising representatives from member states will advise and assist the commission and member states on the consistent and effective application of the AI Act. Additionally, an advisory forum will be formed to provide technical expertise to the AI board and the European Commission.

Compliance

The AI Act stipulates penalties for non-compliance, with fines calculated as a percentage of the offending company’s global annual turnover from the previous financial year or a predetermined amount, whichever is higher. Small and medium-sized enterprises (SMEs) and start-ups will face proportional administrative fines.

Protection of fundamental rights

Before deploying high-risk AI systems in public services, entities must conduct a fundamental rights impact assessment. The regulation mandates increased transparency in the development and use of high-risk AI systems. These systems, along with certain users of high-risk AI, will need to be registered in the EU database for high-risk AI systems. Moreover, users of emotion recognition systems must inform individuals when they are exposed to such technology.

Innovation

The AI Act aims to foster innovation by providing an innovation-friendly legal framework. It promotes evidence-based regulatory learning and establishes AI regulatory sandboxes to enable the controlled development, testing and validation of innovative AI systems under real-world conditions.

Following the signing by the presidents of the European parliament and the council, the AI Act will be published in the EU’s official journal. It will enter into force twenty days after publication and will become applicable two years later, with certain provisions subject to specific exceptions.

Related articles