European Union lawmakers finalized approval for groundbreaking legislation aimed at regulating artificial intelligence (AI), marking a significant milestone in the governance of this rapidly evolving technology.
The Artificial Intelligence Act garnered overwhelming support in the European Parliament in Strasbourg, with 523 votes in favor and 46 against. This legislation, which has been long-anticipated, aims to safeguard citizens across the EU’s 27 member states from potential risks associated with AI while fostering an environment conducive to innovation.
European Parliament President Roberta Metsola hailed the passage of the AI Act as a milestone, emphasizing its potential to position the EU as a global leader in digital and tech innovation rooted in democratic values. Metsola highlighted Europe’s capacity to set international standards responsibly.
Thierry Breton, the EU’s industry chief, expressed satisfaction with the parliament’s strong backing, noting that Europe now emerges as a global benchmark for trustworthy AI.
The new law introduces a framework to categorize different types of AI based on risk levels, imposing stringent requirements or outright bans on AI applications deemed to pose significant threats. While most AI systems are expected to be categorized as low-risk, such as those used for content recommendation or spam filtering, high-risk applications, such as those in the medical sector or critical infrastructure like power networks, will undergo closer scrutiny.
Under the legislation, companies developing high-risk AI models must conduct comprehensive risk assessments, provide transparent information to users, and ensure compliance with regulatory standards before releasing their products to the market. Additionally, the data used to train AI algorithms must meet specified quality and transparency criteria.
The AI Act also includes prohibitions on certain AI applications, including real-time facial recognition in public spaces (with exceptions for law enforcement purposes), predictive policing, and systems that use biometric data to infer an individual’s race, religion, or sexual orientation.
To enforce compliance, companies found violating the law may face fines ranging from €7.5 million to €35 million ($8.2 million to $38.2 million). These penalties underscore the EU’s commitment to upholding rigorous standards and ensuring accountability in the deployment of AI technologies.