EU AI Act Comprehensive; US Relies on State Laws

The European Union has enacted the AI Act, a landmark piece of legislation comprising 113 articles and approximately 50,000 words, establishing the world's first comprehensive regulatory framework for artificial intelligence. In contrast, the United States has not yet passed any federal laws governing AI. Instead, the U.S. approach relies on a decentralized system that includes executive orders, guidance issued by agencies such as the National Institute of Standards and Technology (NIST) and the Federal Trade Commission (FTC), and an increasing number of state-level bills. The volume of AI-related legislation at the state level in the U.S. has seen significant growth, with 131 laws enacted in 2024, more than double the 49 passed in 2023. However, a closer examination of the "weight" and scope of these laws reveals a stark difference in regulatory impact. A single article within the EU AI Act can possess greater legal authority than multiple state statutes in the U.S. This is because U.S. state laws may be narrowly focused and are subject to potential repeal by future federal legislation. Conversely, EU law holds supremacy over the national laws of its member states, meaning a single provision can have far-reaching consequences across entire industries. The EU AI Act distinguishes itself further by outright banning specific categories of AI practices. Article 5 of the Act prohibits eight types of AI applications, including social scoring systems, real-time facial recognition in public spaces for law enforcement purposes, emotion recognition in educational and workplace settings, and predictive policing based solely on profiling. These prohibitions are applied broadly to entire categories deemed unacceptable, although four of these eight banned categories include specific use-case exceptions. For instance, law enforcement agencies are permitted to use facial recognition in emergency situations, such as locating kidnapping victims. The U.S. federal government has not implemented any such category-wide bans on AI technologies. Instead, previous U.S. administrations and subsequent court decisions have tended to focus on regulating specific AI use cases, such as those related to tenant screening or the advertising of housing opportunities, rather than prohibiting the underlying technologies themselves.