Regulation (EU) 2024/1689, known as the AI Act, establishes harmonized rules for artificial intelligence systems within the European Union, aiming to enhance market functioning and ensure safety and fundamental rights. Its extraterritorial scope means it will also apply to providers and deployers from non-EU countries who offer AI systems in the EU.
The European Commission’s Guidelines on the AI Act define an ‘AI system’ as a machine-based entity that operates with varying levels of autonomy and may exhibit adaptiveness after deployment. The Guidelines detail the essential elements of an AI system, including its ability to infer outputs that can influence physical or virtual environments.
The EU is poised to invest hundreds of billions of euros in artificial intelligence to compete with the USA and China, yet risks wasting a significant portion of this funding on compliance with bureaucratic regulations. The situation raises concerns that the AI Act may hinder rather than help the development of competitive AI technologies in Europe.
On February 4, 2025, the European Commission issued draft guidelines clarifying the AI practices prohibited under the EU AI Act, aimed at helping businesses understand their obligations. These guidelines address unacceptable risks such as social scoring, manipulation, and emotion recognition in workplaces and educational institutions.
The EU Artificial Intelligence Act is the world’s first comprehensive regulation for AI, aimed at ensuring responsible AI practices across various sectors. It introduces a risk-based classification system for AI applications, categorizing them from unacceptable risk to minimal risk, each with specific compliance requirements.
The EU has initiated a consultation on the proposed AI Liability Directive (AILD), which aims to establish a liability regime for AI developers and users. This directive seeks to address whether new legislation is necessary and if it should require AI liability insurance.
The European Commission has released comprehensive guidelines defining AI systems under Regulation (EU) 2024/1689, aimed at helping stakeholders determine if their systems qualify as AI. These guidelines emphasize the diverse characteristics of AI, focusing on a lifecycle-based approach that includes both pre-deployment and post-deployment phases.
The EU AI Act, which entered into force on August 1, 2024, establishes a comprehensive framework for the development and use of artificial intelligence within the European Union. It differentiates between AI Systems and General-Purpose AI Models, imposing varying compliance obligations based on the perceived risk level associated with their use.
EU regulators have outlined eight harmful AI practices that will be banned under the new AI Act to ensure safety and prevent abuses of technology. These include mass surveillance, social scoring, and emotion detection in workplaces, aimed at protecting individuals from potential risks associated with AI systems.
From February 2, 2025, AI system providers and deployers must ensure their workforce has adequate AI literacy and eliminate prohibited AI practices. The EU AI Act, effective from August 1, 2024, imposes significant obligations on companies regarding training and compliance with AI regulations.