Answers to Top Questions About the E.U. AI Act: Reach and Unique Requirements

The European Regulation on Artificial Intelligence, or E.U. AI Act (Act), was approved last month and is the world’s first comprehensive AI law. The Act lists extensive mandates for Big AI platforms and providers of high-risk systems but will require all companies with European customers to carefully assess the risks of their AI use. Despite the Act’s 113 chapters and 13 annexes, it lacks details on how to fulfill several crucial mandates. This three-part article series delivers answers to common questions about the Act. Including perspectives from experts at and Orrick, this first installment of the series addresses initial reactions among companies and delves into gray areas around requirements, such as those directed at training data, categorization of providers and deployers, and use of generative AI features. Part two will answer key questions about companies’ risk classifications and requirements based thereon. Part three will offer steps companies can take to prepare for the Act. See “AI Governance Strategies for Privacy Pros” (Apr. 17, 2024).

To read the full article

Continue reading your article with a CSLR subscription.