After years of extensive discussions among member states and members of the European Parliament, a landmark agreement has been reached on regulating Artificial Intelligence (AI) within the European Union. This agreement, hailed as a pioneering step in the digital world, aims to place human interest at the heart of AI legislation.
While the full details of the agreement remain under wraps, European Commissioner Thierry Breton has outlined a two-tier approach. This includes transparency requirements for all general-purpose AI models, like ChatGPT, and stricter mandates for powerful AI systems with systemic impact across the EU. This approach, as reported by the Financial Times, is designed to ensure AI technology is used responsibly while avoiding excessive burdens on companies.
Dragoș Tudorache, a key MEP involved in the AI Act in the Parliament, underscored the significance of this regulation: “We are the first in the world to institute real regulation for AI and for the future AI-driven digital world, guiding the development and evolution of this technology in a human-centric direction.”
Key elements of the new regulation include strict limitations on the use of facial recognition technology, except in clearly defined law enforcement scenarios. The legislation also prohibits AI applications that create a ‘social score’ for individuals or manipulate human behavior to undermine free will. Additionally, the use of AI to exploit vulnerable individuals based on age, disability, or economic status is banned.
Companies failing to adhere to these regulations face hefty penalties, including fines of up to 35 million euros or 7% of their global revenue.
Some of the voices close to the companies already say that the rules seem too restrictive and risk slowing down development. Whether it is so or not, it remains to be seen when a complete version will be published.
