The European Parliament has taken a historic step by adopting the AI Act, which sets out a comprehensive framework of rules and requirements for the development and use of artificial intelligence in Europe. The legislation aims to ensure that AI is safe, transparent, environmentally friendly, and ethically sound.
Risk-Based Approach and Prohibited Practices for AI
The AI Act outlines a risk-based approach to AI, with obligations for developers and users based on the level of risk involved. The use of AI systems that pose an unacceptable level of risk to human safety, such as those using subliminal or targeted manipulative methods, exploiting human vulnerabilities, or used for social assessment, will be strictly prohibited.
The legislation also clarifies a list of practices that are prohibited for artificial intelligence, including remote biometric identification, biometric categorization using sensitive characteristics, predictive policing systems, emotion recognition in law enforcement, border services, workplaces and educational institutions, and indiscriminate extraction of biometric data from social networks or video recordings from CCTV cameras.
“High-Risk” AI Systems and Obligations for Suppliers and Developers
The AI Act specifies “high-risk” AI systems that could potentially harm human health or the environment, including AI systems used to influence voters in political campaigns and social media recommendation algorithms. Suppliers and developers of these “high-risk” AI systems are required by law to conduct risk and compliance assessments, register their systems in the European AI database, ensure high-quality data used for AI training, ensure system transparency and user awareness, as well as mandatory human supervision and the possibility of intervention in the system.
Requirements for “Low-Risk” AI Systems
Providers of “low-risk” AI systems, such as chatbots, must inform users that they are interacting with AI, not a human. For generative models, providers must also inform users that the content is created by a neural network.
Ensuring Safety, Transparency, and Ethical Use of AI
The adoption of the AI Act is a significant step towards ensuring the safety, transparency, and ethical use of AI in Europe. The legislation provides a clear regulatory framework for developers and users of artificial intelligence, while also protecting the fundamental rights of individuals. It is expected to set a precedent for similar legislation in other parts of the world, as the use of AI continues to grow and become an increasingly important part of daily life.
