Brussels

EU restricts AI: The European Parliament this week approved sweeping new rules governing the use of artificial intelligence, beating the U.S. to regulation of the rapidly evolving technology. The EU AI Act divides AI into categories of risk, from “low” to “unacceptable.” Minimally risky uses, like inventory-management software, will require little oversight; unacceptable uses, such as voice-activated children’s toys that encourage dangerous behavior, will be banned. The law also requires manipulated images and videos to be labeled as such. The new rules are expected to be codified into law in May but will take effect gradually over the next few years. They will have global impact far beyond the EU, because international companies, including U.S. tech giants, will have to adapt their AI products for use in the EU market of 448 million people.