The article summarizes the proposed framework of the European Union’s AI Act. First, the article outlines the key features of the EU AI Act including its objectives and scope, the extraterritorial application in scope, a prohibition of certain AI use cases, conditions for AI systems deemed high-risk, the establishment of supervisory authorities, specification of fines, conditional requirements of transparency, and exceptions granted for regulatory sandboxes. The article additionally reviews the required procedure to enact the AI Act, noting both industry’s relief and civil society’s urge for stronger regulation. Ultimately, the author concludes by highlighting that many changes to the proposal are expected to be made, and urges an update addressing unclear and ambiguous language in the proposal.
