CDT Europe’s AI Bulletin for April 2024 reports significant advancements in the regulation of artificial intelligence in Europe. The European Parliament's adoption of the EU AI Act marks a crucial milestone, with 523 MEPs voting in favor, 46 against, and 49 abstentions after two years of intense negotiation.
The AI Act introduces a risk-based regulatory approach, categorizing AI systems based on the level of risk they pose. Eight types of AI systems have been identified as presenting unacceptable risk and are consequently banned. Additionally, high-risk use cases, including critical infrastructure, employment, law enforcement, and migration, are listed in Annex III of the Act.
While the AI Act is hailed as a global precedent for comprehensive AI legislation, concerns have been raised regarding its human rights implications. Despite efforts to enhance transparency and accountability through the creation of a publicly accessible database on high-risk AI, issues persist, particularly in the realm of national security exemptions and certain AI prohibitions losing momentum.
Looking ahead, the AI Act is expected to be endorsed by the Council, published in the Official Journal of the European Union, and enter into force in the coming months. Key provisions will be rolled out in a staggered manner to ensure effective implementation and compliance across member states.
In parallel, the European Commission is launching the AI Pact to promote industry adherence to the AI Act, emphasizing knowledge-sharing and voluntary compliance pledges. Efforts are also underway to appoint AI regulators and establish governance structures to oversee the enforcement of AI regulations at both national and regional levels.
As Europe continues to navigate the complex landscape of AI regulation, the AI Act stands as a significant step towards safeguarding fundamental rights and democracy in the digital age.
