AI Under the European Lens: What the New Law Says

An Innovative Regulatory Framework to Promote Ethical AI Development and Use for the Benefit of Citizens and Businesses

On April 21, 2021, the European Commission presented the AI Act, a proposed regulation for artificial intelligence (AI) aimed at establishing a comprehensive legal framework for its development, market-entry, and use in the European Union. The goal is to promote innovation in the AI sector while ensuring safety, reliability, and the protection of fundamental rights for European citizens. Recently, the European Union approved this significant law to regulate the rapidly growing sector and protect citizens' security and fundamental rights. The law establishes different risk categories for AI systems and imposes specific requirements based on the level of risk.

Main Objectives

1. Safety and Reliability: The AI Act aims to ensure that AI applications are safe and do not cause physical or psychological harm to users, preventing algorithmic biases and discrimination.
2. Transparency and Accountability: Companies and developers must provide clear and transparent information about AI technologies, ensuring users understand how and why certain automated decisions are made.
3. Innovation and Competitiveness: The AI Act promotes innovation in the AI sector by creating a stable regulatory environment that encourages investment and technological development, keeping Europe competitive globally.

Key Points of the AI Law:

• Prohibition of High-Risk AI Systems: Systems that pose an unacceptable threat to the safety, livelihoods, and rights of individuals, such as social scoring by governments and dangerous voice-assisted toys, are prohibited.
• Stringent Requirements for High-Risk AI Systems: Systems used in sectors such as transportation, education, credit management, and law enforcement must meet strict requirements, including risk assessments and mitigation measures, before being placed on the market.
• Transparency Obligations for Limited-Risk AI Systems: Systems like chatbots and spam filters must be transparent about AI use, informing users and allowing them to choose whether to continue the interaction.
• Development of Reliable AI Models: Specific standards for large AI models aim to ensure responsible development and use of these models, adequately managing risks.
• Governance and Enforcement: A new European Artificial Intelligence Office will oversee the law's implementation, working with Member States to ensure its consistent application across the EU.

AI System Classification

1. Unacceptable Risk: AI systems that pose a clear threat to safety, fundamental rights, or European values, such as social scoring systems, are banned.
2. High Risk: AI applications in critical sectors such as health, education, and justice are subject to strict controls and must meet specific safety and transparency requirements.
3. Limited Risk: AI systems that require transparency obligations, such as chatbots, must inform users that they are interacting with artificial intelligence.
4. Minimal Risk: Most AI applications, which pose low risk, are not subject to additional specific requirements.

Risk-Based Requirements

AI system providers must comply with specific requirements based on the risk category of their system. The requirements for high-risk systems are the most stringent and include:

• Conformity Assessment: Providers must evaluate their system by an independent body before it can be placed on the market.
• Transparency Obligations: Providers must provide clear and understandable information about their system, including its purpose, capabilities, and limitations.
• Safety Measures: Providers must implement adequate safety measures to protect their systems from cyber-attacks and other risks.
• Risk Management Systems: Providers must implement risk management systems to identify, assess, and mitigate the risks associated with their systems.

Impact on the Industry

The AI Act will significantly impact the European and global technology industry. Companies must adapt their development and implementation processes to comply with the new standards, potentially increasing compliance costs and improving user trust in AI technologies.

A "Future-Proof" Approach

The AI law is designed to be "future-proof," meaning it can adapt to technological advances. The European Commission will have the power to update the law as necessary to ensure it protects citizens' safety and fundamental rights.

Impact of the AI Law

The AI Act is a significant step forward for the responsible development of artificial intelligence. The law will help ensure that AI systems are used safely, ethically, and in a manner that respects citizens' fundamental rights. Additionally, the law will help position Europe as a global leader in developing reliable AI technologies.

The AI Law is expected to come into effect in 2026, with some exceptions for high-risk systems. The European Commission has already launched the AI Pact, a voluntary initiative to help businesses comply with the law's requirements.

Conclusion

The approval of the AI Act represents an important step for the European Union. The law will help ensure that artificial intelligence is developed and used safely and responsibly, benefiting everyone.

Share